markdown
stringlengths 0
1.02M
| code
stringlengths 0
832k
| output
stringlengths 0
1.02M
| license
stringlengths 3
36
| path
stringlengths 6
265
| repo_name
stringlengths 6
127
|
|---|---|---|---|---|---|
Hyperparameter Tuning in KNN Manually finding the optimal value of n_neighbors parameter | # Find the optimal value of the n_neighbors parameter
models={f'KNN_{i}':KNeighborsClassifier(n_neighbors=i) for i in range(2,31)}
# run the model only for fold number 4 ie the 5th fold
accuracy,confusion_matrices,classification_report=run(fold=4,df=df_optimal_KNN,models=models,print_details=True)
x=[i for i in range(2,31)]
y=accuracy
plt.plot(x,y)
plt.xlabel('Number of Nearest Neighbors')
plt.ylabel('Accuracy Score')
plt.title("Optimal n_neighbors value")
plt.show() | _____no_output_____ | MIT | Diabetes.ipynb | AryanMethil/Diabetes-KNN-vs-Naive-Bayes |
Using Grid Search to find optimal values of n_neighbors and p | from sklearn import model_selection
from sklearn import metrics
def hyperparameter_tune_and_run(df,num_folds,models,target_name,param_grid,evaluation_metric,print_details=False):
X=df.drop(labels=[target_name,'kfolds'],axis=1).values
y=df[target_name]
model_name,model_constructor=list(models.items())[0]
model = model_selection.GridSearchCV(
estimator = model_constructor,
param_grid = param_grid,
scoring = evaluation_metric,
verbose = 10,
cv = num_folds
)
model.fit(X,y)
if(print_details==True):
print(f"Best score : {model.best_score_}")
print("Best parameters : ")
best_parameters=model.best_estimator_.get_params()
for param_name in sorted(param_grid.keys()):
print(f"\t{param_name}: {best_parameters[param_name]}")
return model
models={'KNN': KNeighborsClassifier()}
param_grid = {
"n_neighbors" : [i for i in range(2,31)],
"p" : [2,3]
}
model = hyperparameter_tune_and_run(df=df_optimal_KNN,num_folds=5,models=models,target_name='Outcome',param_grid=param_grid,evaluation_metric="accuracy",print_details=True) | Fitting 5 folds for each of 58 candidates, totalling 290 fits
[CV] n_neighbors=2, p=2 ..............................................
[CV] .................. n_neighbors=2, p=2, score=0.770, total= 0.0s
[CV] n_neighbors=2, p=2 ..............................................
[CV] .................. n_neighbors=2, p=2, score=0.730, total= 0.0s
[CV] n_neighbors=2, p=2 ..............................................
[CV] .................. n_neighbors=2, p=2, score=0.785, total= 0.0s
[CV] n_neighbors=2, p=2 ..............................................
[CV] .................. n_neighbors=2, p=2, score=0.755, total= 0.0s
[CV] n_neighbors=2, p=2 ..............................................
[CV] .................. n_neighbors=2, p=2, score=0.740, total= 0.0s
[CV] n_neighbors=2, p=3 ..............................................
[CV] .................. n_neighbors=2, p=3, score=0.775, total= 0.0s
[CV] n_neighbors=2, p=3 ..............................................
[CV] .................. n_neighbors=2, p=3, score=0.725, total= 0.0s
[CV] n_neighbors=2, p=3 ..............................................
[CV] .................. n_neighbors=2, p=3, score=0.790, total= 0.0s
[CV] n_neighbors=2, p=3 ..............................................
[CV] .................. n_neighbors=2, p=3, score=0.745, total= 0.0s
[CV] n_neighbors=2, p=3 ..............................................
[CV] .................. n_neighbors=2, p=3, score=0.745, total= 0.0s
[CV] n_neighbors=3, p=2 ..............................................
[CV] .................. n_neighbors=3, p=2, score=0.750, total= 0.0s
[CV] n_neighbors=3, p=2 ..............................................
[CV] .................. n_neighbors=3, p=2, score=0.750, total= 0.0s
[CV] n_neighbors=3, p=2 ..............................................
[CV] .................. n_neighbors=3, p=2, score=0.740, total= 0.0s
[CV] n_neighbors=3, p=2 ..............................................
[CV] .................. n_neighbors=3, p=2, score=0.755, total= 0.0s
[CV] n_neighbors=3, p=2 ..............................................
[CV] .................. n_neighbors=3, p=2, score=0.745, total= 0.0s
[CV] n_neighbors=3, p=3 ..............................................
[CV] .................. n_neighbors=3, p=3, score=0.745, total= 0.0s
[CV] n_neighbors=3, p=3 ..............................................
| MIT | Diabetes.ipynb | AryanMethil/Diabetes-KNN-vs-Naive-Bayes |
Comparison between KNN and NB1. Dataset when KNN was considered for feature selection2. Dataset when NB was considered for feature selection | # Compare between KNN and Naive Bayes
models={
'KNN': KNeighborsClassifier(n_neighbors=12,p=3),
'Gaussian Naive Bayes': GaussianNB(),
}
# accuracies => list of 5 lists. Each list will contain 3 values ie KNN accuracy, Gaussian Naive Bayes
accuracies,confusion_matrices,classification_reports=[],[],[]
for f in range(5):
accuracy,confusion_matrix,classification_report=run(f,df_optimal_KNN,models=models,print_details=True)
accuracies.append(accuracy)
confusion_matrices.append(confusion_matrix)
classification_reports.append(classification_report)
print(accuracies)
x_axis_labels=['Predicted Normal','Predicted Diabetic']
y_axis_labels=['True Normal','True Diabetic']
import seaborn as sns
# Heatmap of confusion matrix of 5th fold of KNN
sns.heatmap(confusion_matrices[4][0],xticklabels=x_axis_labels,yticklabels=y_axis_labels,annot=True)
# Heatmap of confusion matrix of 5th fold of Naive Bayes
sns.heatmap(confusion_matrices[4][1],xticklabels=x_axis_labels,yticklabels=y_axis_labels,annot=True)
# Classification report of 5th fold of KNN
print("KNN")
print(classification_reports[4][0])
# Classification report of 5th fold of Naive Bayes
print("Naive Bayes")
print(classification_reports[4][1])
accuracies,confusion_matrices,classification_reports=[],[],[]
for f in range(5):
accuracy,confusion_matrix,classification_report=run(f,df_optimal_NB,models=models,print_details=True)
accuracies.append(accuracy)
confusion_matrices.append(confusion_matrix)
classification_reports.append(classification_report)
import seaborn as sns
# Heatmap of confusion matrix of 5th fold of KNN
sns.heatmap(confusion_matrices[4][0],xticklabels=x_axis_labels,yticklabels=y_axis_labels,annot=True)
# Heatmap of confusion matrix of 5th fold of Naive Bayes
sns.heatmap(confusion_matrices[4][1],xticklabels=x_axis_labels,yticklabels=y_axis_labels,annot=True)
# Classification report of 5th fold of KNN
print("KNN")
print(classification_reports[4][0])
# Classification report of 5th fold of Naive Bayes
print("Naive Bayes")
print(classification_reports[4][1])
| _____no_output_____ | MIT | Diabetes.ipynb | AryanMethil/Diabetes-KNN-vs-Naive-Bayes |
Getting info on Priming experiment dataset that's needed for modeling Info:* __Which gradient(s) to simulate?__* For each gradient to simulate: * Infer total richness of starting community * Get distribution of total OTU abundances per fraction * Number of sequences per sample * Infer total abundance of each target taxon User variables | baseDir = '/home/nick/notebook/SIPSim/dev/priming_exp/'
workDir = os.path.join(baseDir, 'exp_info')
otuTableFile = '/var/seq_data/priming_exp/data/otu_table.txt'
otuTableSumFile = '/var/seq_data/priming_exp/data/otu_table_summary.txt'
metaDataFile = '/var/seq_data/priming_exp/data/allsample_metadata_nomock.txt'
#otuRepFile = '/var/seq_data/priming_exp/otusn.pick.fasta'
#otuTaxFile = '/var/seq_data/priming_exp/otusn_tax/otusn_tax_assignments.txt'
#genomeDir = '/home/nick/notebook/SIPSim/dev/bac_genome1210/genomes/' | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
Init | import glob
%load_ext rpy2.ipython
%%R
library(ggplot2)
library(dplyr)
library(tidyr)
library(gridExtra)
library(fitdistrplus)
if not os.path.isdir(workDir):
os.makedirs(workDir) | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
Loading OTU table (filter to just bulk samples) | %%R -i otuTableFile
tbl = read.delim(otuTableFile, sep='\t')
# filter
tbl = tbl %>%
select(ends_with('.NA'))
tbl %>% ncol %>% print
tbl[1:4,1:4]
%%R
tbl.h = tbl %>%
gather('sample', 'count', 1:ncol(tbl)) %>%
separate(sample, c('isotope','treatment','day','rep','fraction'), sep='\\.', remove=F)
tbl.h %>% head | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
Which gradient(s) to simulate? | %%R -w 900 -h 400
tbl.h.s = tbl.h %>%
group_by(sample) %>%
summarize(total_count = sum(count)) %>%
separate(sample, c('isotope','treatment','day','rep','fraction'), sep='\\.', remove=F)
ggplot(tbl.h.s, aes(day, total_count, color=rep %>% as.character)) +
geom_point() +
facet_grid(isotope ~ treatment) +
theme(
text = element_text(size=16)
)
%%R
tbl.h.s$sample[grepl('700', tbl.h.s$sample)] %>% as.vector %>% sort | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
NotesSamples to simulate* Isotope: * 12C vs 13C* Treatment: * 700* Days: * 14 * 28 * 45 | %%R
# bulk soil samples for gradients to simulate
samples.to.use = c(
"X12C.700.14.05.NA",
"X12C.700.28.03.NA",
"X12C.700.45.01.NA",
"X13C.700.14.08.NA",
"X13C.700.28.06.NA",
"X13C.700.45.01.NA"
) | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
Total richness of starting (bulk-soil) communityMethod:* Total number of OTUs in OTU table (i.e., gamma richness)* Just looking at bulk soil samples Loading just bulk soil | %%R -i otuTableFile
tbl = read.delim(otuTableFile, sep='\t')
# filter
tbl = tbl %>%
select(ends_with('.NA'))
tbl$OTUId = rownames(tbl)
tbl %>% ncol %>% print
tbl[1:4,1:4]
%%R
tbl.h = tbl %>%
gather('sample', 'count', 1:(ncol(tbl)-1)) %>%
separate(sample, c('isotope','treatment','day','rep','fraction'), sep='\\.', remove=F)
tbl.h %>% head
%%R -w 800
tbl.s = tbl.h %>%
filter(count > 0) %>%
group_by(sample, isotope, treatment, day, rep, fraction) %>%
summarize(n_taxa = n())
ggplot(tbl.s, aes(day, n_taxa, color=rep %>% as.character)) +
geom_point() +
facet_grid(isotope ~ treatment) +
theme_bw() +
theme(
text = element_text(size=16),
axis.text.x = element_blank()
)
%%R -w 800 -h 350
# filter to just target samples
tbl.s.f = tbl.s %>% filter(sample %in% samples.to.use)
ggplot(tbl.s.f, aes(day, n_taxa, fill=rep %>% as.character)) +
geom_bar(stat='identity') +
facet_grid(. ~ isotope) +
labs(y = 'Number of taxa') +
theme_bw() +
theme(
text = element_text(size=16),
axis.text.x = element_blank()
)
%%R
message('Bulk soil total observed richness: ')
tbl.s.f %>% select(-fraction) %>% as.data.frame %>% print | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
Number of taxa in all fractions corresponding to each bulk soil sample* Trying to see the difference between richness of bulk vs gradients (veil line effect) | %%R -i otuTableFile
# loading OTU table
tbl = read.delim(otuTableFile, sep='\t') %>%
select(-ends_with('.NA'))
tbl.h = tbl %>%
gather('sample', 'count', 2:ncol(tbl)) %>%
separate(sample, c('isotope','treatment','day','rep','fraction'), sep='\\.', remove=F)
tbl.h %>% head
%%R
# basename of fractions
samples.to.use.base = gsub('\\.[0-9]+\\.NA', '', samples.to.use)
samps = tbl.h$sample %>% unique
fracs = sapply(samples.to.use.base, function(x) grep(x, samps, value=TRUE))
for (n in names(fracs)){
n.frac = length(fracs[[n]])
cat(n, '-->', 'Number of fraction samples: ', n.frac, '\n')
}
%%R
# function for getting all OTUs in a sample
n.OTUs = function(samples, otu.long){
otu.long.f = otu.long %>%
filter(sample %in% samples,
count > 0)
n.OTUs = otu.long.f$OTUId %>% unique %>% length
return(n.OTUs)
}
num.OTUs = lapply(fracs, n.OTUs, otu.long=tbl.h)
num.OTUs = do.call(rbind, num.OTUs) %>% as.data.frame
colnames(num.OTUs) = c('n_taxa')
num.OTUs$sample = rownames(num.OTUs)
num.OTUs
%%R
tbl.s.f %>% as.data.frame
%%R
# joining with bulk soil sample summary table
num.OTUs$data = 'fractions'
tbl.s.f$data = 'bulk_soil'
tbl.j = rbind(num.OTUs,
tbl.s.f %>% ungroup %>% select(sample, n_taxa, data)) %>%
mutate(isotope = gsub('X|\\..+', '', sample),
sample = gsub('\\.[0-9]+\\.NA', '', sample))
tbl.j
%%R -h 300 -w 800
ggplot(tbl.j, aes(sample, n_taxa, fill=data)) +
geom_bar(stat='identity', position='dodge') +
facet_grid(. ~ isotope, scales='free_x') +
labs(y = 'Number of OTUs') +
theme(
text = element_text(size=16)
# axis.text.x = element_text(angle=90)
) | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
Distribution of total sequences per fraction * Number of sequences per sample* Using all samples to assess this one* Just fraction samples__Method:__* Total number of sequences (total abundance) per sample Loading OTU table | %%R -i otuTableFile
tbl = read.delim(otuTableFile, sep='\t')
# filter
tbl = tbl %>%
select(-ends_with('.NA'))
tbl %>% ncol %>% print
tbl[1:4,1:4]
%%R
tbl.h = tbl %>%
gather('sample', 'count', 2:ncol(tbl)) %>%
separate(sample, c('isotope','treatment','day','rep','fraction'), sep='\\.', remove=F)
tbl.h %>% head
%%R -h 400
tbl.h.s = tbl.h %>%
group_by(sample) %>%
summarize(total_seqs = sum(count))
p = ggplot(tbl.h.s, aes(total_seqs)) +
theme_bw() +
theme(
text = element_text(size=16)
)
p1 = p + geom_histogram(binwidth=200)
p2 = p + geom_density()
grid.arrange(p1,p2,ncol=1) | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
Distribution fitting | %%R -w 700 -h 350
plotdist(tbl.h.s$total_seqs)
%%R -w 450 -h 400
descdist(tbl.h.s$total_seqs, boot=1000)
%%R
f.n = fitdist(tbl.h.s$total_seqs, 'norm')
f.ln = fitdist(tbl.h.s$total_seqs, 'lnorm')
f.ll = fitdist(tbl.h.s$total_seqs, 'logis')
#f.c = fitdist(tbl.s$count, 'cauchy')
f.list = list(f.n, f.ln, f.ll)
plot.legend = c('normal', 'log-normal', 'logistic')
par(mfrow = c(2,1))
denscomp(f.list, legendtext=plot.legend)
qqcomp(f.list, legendtext=plot.legend)
%%R
gofstat(list(f.n, f.ln, f.ll), fitnames=plot.legend)
%%R
summary(f.ln) | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
Notes:* best fit: * lognormal * mean = 10.113 * sd = 1.192 Does sample size correlate to buoyant density? Loading OTU table | %%R -i otuTableFile
tbl = read.delim(otuTableFile, sep='\t')
# filter
tbl = tbl %>%
select(-ends_with('.NA')) %>%
select(-starts_with('X0MC'))
tbl = tbl %>%
gather('sample', 'count', 2:ncol(tbl)) %>%
mutate(sample = gsub('^X', '', sample))
tbl %>% head
%%R
# summarize
tbl.s = tbl %>%
group_by(sample) %>%
summarize(total_count = sum(count))
tbl.s %>% head(n=3) | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
Loading metadata | %%R -i metaDataFile
tbl.meta = read.delim(metaDataFile, sep='\t')
tbl.meta %>% head(n=3) | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
Determining association | %%R -w 700
tbl.j = inner_join(tbl.s, tbl.meta, c('sample' = 'Sample'))
ggplot(tbl.j, aes(Density, total_count, color=rep)) +
geom_point() +
facet_grid(Treatment ~ Day)
%%R -w 600 -h 350
ggplot(tbl.j, aes(Density, total_count)) +
geom_point(aes(color=Treatment)) +
geom_smooth(method='lm') +
labs(x='Buoyant density', y='Total sequences') +
theme_bw() +
theme(
text = element_text(size=16)
) | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
Number of taxa along the gradient | %%R
tbl.s = tbl %>%
filter(count > 0) %>%
group_by(sample) %>%
summarize(n_taxa = sum(count > 0))
tbl.j = inner_join(tbl.s, tbl.meta, c('sample' = 'Sample'))
tbl.j %>% head(n=3)
%%R -w 900 -h 600
ggplot(tbl.j, aes(Density, n_taxa, fill=rep, color=rep)) +
#geom_area(stat='identity', alpha=0.5, position='dodge') +
geom_point() +
geom_line() +
labs(x='Buoyant density', y='Number of taxa') +
facet_grid(Treatment ~ Day) +
theme_bw() +
theme(
text = element_text(size=16),
legend.position = 'none'
) | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
Notes:* Many taxa out to the tails of the gradient.* It seems that the DNA fragments were quite diffuse in the gradients. Total abundance of each target taxon: bulk soil approach* Getting relative abundances from bulk soil samples * This has the caveat of likely undersampling richness vs using all gradient fraction samples. * i.e., veil line effect | %%R -i otuTableFile
# loading OTU table
tbl = read.delim(otuTableFile, sep='\t')
# filter
tbl = tbl %>%
select(matches('OTUId'), ends_with('.NA'))
tbl %>% ncol %>% print
tbl[1:4,1:4]
%%R
# long table format w/ selecting samples of interest
tbl.h = tbl %>%
gather('sample', 'count', 2:ncol(tbl)) %>%
separate(sample, c('isotope','treatment','day','rep','fraction'), sep='\\.', remove=F) %>%
filter(sample %in% samples.to.use,
count > 0)
tbl.h %>% head
%%R
message('Number of samples: ', tbl.h$sample %>% unique %>% length)
message('Number of OTUs: ', tbl.h$OTUId %>% unique %>% length)
%%R
tbl.hs = tbl.h %>%
group_by(OTUId) %>%
summarize(
total_count = sum(count),
mean_count = mean(count),
median_count = median(count),
sd_count = sd(count)
) %>%
filter(total_count > 0)
tbl.hs %>% head | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
For each sample, writing a table of OTU_ID and count | %%R -i workDir
setwd(workDir)
samps = tbl.h$sample %>% unique %>% as.vector
for(samp in samps){
outFile = paste(c(samp, 'OTU.txt'), collapse='_')
tbl.p = tbl.h %>%
filter(sample == samp, count > 0)
write.table(tbl.p, outFile, sep='\t', quote=F, row.names=F)
message('Table written: ', outFile)
message(' Number of OTUs: ', tbl.p %>% nrow)
} | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
Making directories for simulations | p = os.path.join(workDir, '*_OTU.txt')
files = glob.glob(p)
baseDir = os.path.split(workDir)[0]
newDirs = [os.path.split(x)[1].rstrip('.NA_OTU.txt') for x in files]
newDirs = [os.path.join(baseDir, x) for x in newDirs]
for newDir,f in zip(newDirs, files):
if not os.path.isdir(newDir):
print 'Making new directory: {}'.format(newDir)
os.makedirs(newDir)
else:
print 'Directory exists: {}'.format(newDir)
# symlinking file
linkPath = os.path.join(newDir, os.path.split(f)[1])
if not os.path.islink(linkPath):
os.symlink(f, linkPath) | Directory exists: /home/nick/notebook/SIPSim/dev/priming_exp/X13C.700.28.06
Directory exists: /home/nick/notebook/SIPSim/dev/priming_exp/X12C.700.28.03
Directory exists: /home/nick/notebook/SIPSim/dev/priming_exp/X13C.700.14.08
Directory exists: /home/nick/notebook/SIPSim/dev/priming_exp/X13C.700.45.01
Directory exists: /home/nick/notebook/SIPSim/dev/priming_exp/X12C.700.45.01
Directory exists: /home/nick/notebook/SIPSim/dev/priming_exp/X12C.700.14.05
| MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
Rank-abundance distribution for each sample | %%R -i otuTableFile
tbl = read.delim(otuTableFile, sep='\t')
# filter
tbl = tbl %>%
select(matches('OTUId'), ends_with('.NA'))
tbl %>% ncol %>% print
tbl[1:4,1:4]
%%R
# long table format w/ selecting samples of interest
tbl.h = tbl %>%
gather('sample', 'count', 2:ncol(tbl)) %>%
separate(sample, c('isotope','treatment','day','rep','fraction'), sep='\\.', remove=F) %>%
filter(sample %in% samples.to.use,
count > 0)
tbl.h %>% head
%%R
# ranks of relative abundances
tbl.r = tbl.h %>%
group_by(sample) %>%
mutate(perc_rel_abund = count / sum(count) * 100,
rank = row_number(-perc_rel_abund)) %>%
unite(day_rep, day, rep, sep='-')
tbl.r %>% as.data.frame %>% head(n=3)
%%R -w 900 -h 350
ggplot(tbl.r, aes(rank, perc_rel_abund)) +
geom_point() +
# labs(x='Buoyant density', y='Number of taxa') +
facet_wrap(~ day_rep) +
theme_bw() +
theme(
text = element_text(size=16),
legend.position = 'none'
) | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
Taxon abundance range for each sample-fraction | %%R -i otuTableFile
tbl = read.delim(otuTableFile, sep='\t')
# filter
tbl = tbl %>%
select(-ends_with('.NA')) %>%
select(-starts_with('X0MC'))
tbl = tbl %>%
gather('sample', 'count', 2:ncol(tbl)) %>%
mutate(sample = gsub('^X', '', sample))
tbl %>% head
%%R
tbl.ar = tbl %>%
#mutate(fraction = gsub('.+\\.', '', sample) %>% as.numeric) %>%
#mutate(treatment = gsub('(.+)\\..+', '\\1', sample)) %>%
group_by(sample) %>%
mutate(rel_abund = count / sum(count)) %>%
summarize(abund_range = max(rel_abund) - min(rel_abund)) %>%
ungroup() %>%
separate(sample, c('isotope','treatment','day','rep','fraction'), sep='\\.', remove=F)
tbl.ar %>% head(n=3)
%%R -w 800
tbl.ar = tbl.ar %>%
mutate(fraction = as.numeric(fraction))
ggplot(tbl.ar, aes(fraction, abund_range, fill=rep, color=rep)) +
geom_point() +
geom_line() +
labs(x='Buoyant density', y='Range of relative abundance values') +
facet_grid(treatment ~ day) +
theme_bw() +
theme(
text = element_text(size=16),
legend.position = 'none'
) | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
Total abundance of each target taxon: all fraction samples approach* Getting relative abundances from all fraction samples for the gradient * I will need to calculate (mean|max?) relative abundances for each taxon and then re-scale so that cumsum = 1 | %%R -i otuTableFile
# loading OTU table
tbl = read.delim(otuTableFile, sep='\t') %>%
select(-ends_with('.NA'))
tbl.h = tbl %>%
gather('sample', 'count', 2:ncol(tbl)) %>%
separate(sample, c('isotope','treatment','day','rep','fraction'), sep='\\.', remove=F)
tbl.h %>% head
%%R
# basename of fractions
samples.to.use.base = gsub('\\.[0-9]+\\.NA', '', samples.to.use)
samps = tbl.h$sample %>% unique
fracs = sapply(samples.to.use.base, function(x) grep(x, samps, value=TRUE))
for (n in names(fracs)){
n.frac = length(fracs[[n]])
cat(n, '-->', 'Number of fraction samples: ', n.frac, '\n')
}
%%R
# function for getting mean OTU abundance from all fractions
OTU.abund = function(samples, otu.long){
otu.rel.abund = otu.long %>%
filter(sample %in% samples,
count > 0) %>%
ungroup() %>%
group_by(sample) %>%
mutate(total_count = sum(count)) %>%
ungroup() %>%
mutate(perc_abund = count / total_count * 100) %>%
group_by(OTUId) %>%
summarize(mean_perc_abund = mean(perc_abund),
median_perc_abund = median(perc_abund),
max_perc_abund = max(perc_abund))
return(otu.rel.abund)
}
## calling function
otu.rel.abund = lapply(fracs, OTU.abund, otu.long=tbl.h)
otu.rel.abund = do.call(rbind, otu.rel.abund) %>% as.data.frame
otu.rel.abund$sample = gsub('\\.[0-9]+$', '', rownames(otu.rel.abund))
otu.rel.abund %>% head
%%R -h 600 -w 900
# plotting
otu.rel.abund.l = otu.rel.abund %>%
gather('abund_stat', 'value', mean_perc_abund, median_perc_abund, max_perc_abund)
otu.rel.abund.l$OTUId = reorder(otu.rel.abund.l$OTUId, -otu.rel.abund.l$value)
ggplot(otu.rel.abund.l, aes(OTUId, value, color=abund_stat)) +
geom_point(shape='O', alpha=0.7) +
scale_y_log10() +
facet_grid(abund_stat ~ sample) +
theme_bw() +
theme(
text = element_text(size=16),
axis.text.x = element_blank(),
legend.position = 'none'
) | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
For each sample, writing a table of OTU_ID and count | %%R -i workDir
setwd(workDir)
# each sample is a file
samps = otu.rel.abund.l$sample %>% unique %>% as.vector
for(samp in samps){
outFile = paste(c(samp, 'frac_OTU.txt'), collapse='_')
tbl.p = otu.rel.abund %>%
filter(sample == samp, mean_perc_abund > 0)
write.table(tbl.p, outFile, sep='\t', quote=F, row.names=F)
cat('Table written: ', outFile, '\n')
cat(' Number of OTUs: ', tbl.p %>% nrow, '\n')
} | _____no_output_____ | MIT | ipynb/bac_genome/priming_exp/priming_exp_info.ipynb | arischwartz/test |
Here Z ⼠binomial(1, 0.5) is the protected attribute. Features related to the protected attribute are sampled from X ⼠N(¡, I) with ¡ = 1 when Z = 0 and ¡ = 2 when Z = 1. Other features not related to the protected attribute Z are generated with ¡ = 0. First 4 features are correlated with z. The first 10 features are correlated with y according to a logistic regression model y = logit(β^TX) with β ⼠N(¡β, 0.1), where ¡β = 5 for the first 6 features and ¡β = 0 for all others. | z = np.zeros(1000)
for j in range(1000):
z[j] = np.random.binomial(1,0.5)
x_correlated = np.zeros((1000,4))
x_uncorrelated = np.zeros((1000,16))
for j in range(16):
for i in range (1000):
if j < 4:
x_correlated[i][j] = np.random.normal((z[i]*2 + 10), 1, 1)
x_uncorrelated[i][j] = np.random.normal(0,1,1)
x = np.concatenate((x_correlated,x_uncorrelated),axis=1)
x = np.concatenate((x,np.reshape(z,(1000,1))),axis=1)
b = np.zeros(21)
noise = np.random.normal(0,1,1000)
for i in range (10):
b[i] = np.random.normal(5,0.1,1)
y = logit(NormalizeData(np.dot(x,b)) + noise.T)
for i in range (len(y)):
if y[i] > 0:
y[i] = int(1)
else:
y[i] = int(0)
column = []
for i in range(21):
column.append(str(i+1))
dataframe = pd.DataFrame(x, columns = column)
model_dtree = d_tree.DecisionTree(20,0,'21',1)
model_dtree.fit(dataframe,y)
fairness_importance = model_dtree._fairness_importance()
feature = []
score = []
for key, value in fairness_importance.items():
print(key, value)
feature.append(key)
score.append((value))
utils.draw_plot(feature,score,"Results/Synthetic/eqop.pdf")
model_dtree_dp = d_tree.DecisionTree(20,0,'21',2)
model_dtree_dp.fit(dataframe,y)
fairness_importance_dp = model_dtree_dp._fairness_importance()
feature = []
score_dp = []
for key, value in fairness_importance_dp.items():
print(key, value)
feature.append(key)
score_dp.append((value))
utils.draw_plot(feature,score_dp,"Results/Synthetic/DP.pdf")
count_z0 = count_z1 = 0
count0 = count1 = 0
z0 = z1 = 0
for i in range (1000):
if y[i] == 0:
count0+=1
else:
count1+=1
if x[i][20] == 0:
count_z0 += 1
else:
count_z1 +=1
if x[i][20] == 0:
z0+=1
else:
z1+=1
print(count0,count1, count_z0,count_z1,z0,z1)
| 809 191 104 87 498 502
| MIT | benchmark/synthetic.ipynb | DebolinaHalder/599 |
 Annif tutorial with Jupyter notebook [Annif](https://annif.org/) is an open source subject indexing tool for new documents and aims to improve the discoverability of vast amount of electronic documents. In order to accomplish automatic subject indexing task, annif uses ML/NLP algorithms to leverage existing training data in the form of subject vocabulary and metadata. For the purpose of a test use-case service of Annif at CSC, small subsets of yso-finna-theses records (i.e., yso-finna-small.tsv.gz file as provided by Annif tutorial dataset) from the [Finna.fi](https://www.finna.fi/?lng=en-gb) discovery serviceis are used. The backend models of this annif instance inculdes handful of different subject indexing algorithms namely, Maui, TF-IDF, ensemble and Omikuji (Parabel/Bonsai) methods. These models are trained in supercomputing (Puhti) environment using singularity container for Annif application. This tutorial uses REST API calls to interact with a given Annif webserver which can be either [Annif webserver](https://api.annif.org/v1/ui/) hosted by national library of Finland or a [test case webserver](https://annif.rahtiapp.fi/v1/ui/) hosted by CSC Learning ObjectivesUpon completion of this tutorial, you will be able to learn how to: - List available trained projects in a given annif webserver - Perform subject indexing with Annif using different projects (subject vocabularies and existing metadata) List all available projects from a given annif webserverAll available projects form annif webserver can be retrieved using Annif REST API GET call. >**Note**: One can make REST API call using the following *curl* command in command-line environment: curl -X GET --header 'Accept: application/json' 'https://annif.rahtiapp.fi/v1/projects' Below is the python way of making REST API GET call to annif server and then converting the resulting json data in the form of a table | import requests
import json
from pandas import json_normalize
headers = {'Accept': 'application/json'}
base_url='https://annif.rahtiapp.fi/v1/projects' # Annif webserver hosted by CSC
#base_url='https://api.annif.org/v1/projects' # Annif webserver by NatLibFi
response = requests.get(base_url, headers=headers)
d=response.json() # print resulting json file: print(response.text)
json_normalize(d['projects']) | _____no_output_____ | Apache-2.0 | annif.ipynb | CSCfi/annif-utils |
Perform subject indexing with AnnifThere are mainly six types of projects in a test case example of [Annif](https://annif.rahtiapp.fi) hosted at CSC and it runson Rahti container cloud. Let's see how to get subject indexing with each of these projects here 1. Perform subject indexing for your own text using YSO TFIDF project (projectid: yso-tfidf-en) This can be accomplished using swagger API POST call> **Note**: Here is the curl command for subject indexing: curl -X POST --header 'Content-Type: application/x-www-form-urlencoded' --header 'Accept: application/json' -d 'text=frequently occurring or otherwise salient terms in the document are matched with terms in the vocabulary&limit=10' 'https://annif.rahtiapp.fi/v1/projects/yso-tfidf-en/suggest'Below is the python approach: | projectid='yso-tfidf-en'
text='frequently occurring or otherwise salient terms in the document are matched with terms in the vocabulary'
url= base_url+ '/' + projectid +'/suggest'
data = {'text': text}
headers = {'Content-Type': 'application/x-www-form-urlencoded','Accept': 'application/json'}
response = requests.post(url, headers=headers, data=data)
d=response.json() # print(response.text)
display(json_normalize(d['results']))
data=json_normalize(d['results'])
data.loc[:,['label','score','uri']].plot('label',kind='bar') | _____no_output_____ | Apache-2.0 | annif.ipynb | CSCfi/annif-utils |
2. Perform subject indexing with YSO ensemble project (projectid:'yso-ensemble-en') This can be accomplished using swagger API POST call> **Note**: curl command - curl -X POST --header 'Content-Type: application/x-www-form-urlencoded' --header 'Accept: application/json' -d 'text=frequently occurring or otherwise salient terms in the document are matched with terms in the vocabulary&limit=10' 'https://annif.rahtiapp.fi/v1/projects/yso-tfidf-en/suggest'Using python: | projectid='yso-ensemble-en'
text='frequently occurring or otherwise salient terms in the document are matched with terms in the vocabulary'
url= base_url+ '/' + projectid +'/suggest'
data = {'text': text}
headers = {'Content-Type': 'application/x-www-form-urlencoded','Accept': 'application/json'}
response = requests.post(url, headers=headers, data=data)
d=response.json() # print(response.text)
display(json_normalize(d['results']))
data=json_normalize(d['results'])
data.loc[:,['label','score','uri']].plot('label',kind='bar') | _____no_output_____ | Apache-2.0 | annif.ipynb | CSCfi/annif-utils |
3. Perform subject indexing with for your own text using project 'yso-maui-en' This can be accomplished using swagger API POST call> **curl command**:curl -X POST --header 'Content-Type: application/x-www-form-urlencoded' --header 'Accept: application/json' -d 'text=frequently occurring or otherwise salient terms in the document are matched with terms in the vocabulary&limit=10' 'https://annif.rahtiapp.fi/v1/projects/yso-maui-en/suggest' Using python approach: | projectid='yso-maui-en'
text='frequently occurring or otherwise salient terms in the document are matched with terms in the vocabulary'
url= base_url+ '/' + projectid +'/suggest'
data = {'text': text}
headers = { 'Content-Type': 'application/x-www-form-urlencoded','Accept': 'application/json'}
response = requests.post(url, headers=headers, data=data)
d=response.json() # print(response.text)
display(json_normalize(d['results']))
data=json_normalize(d['results'])
data.loc[:,['label','score','uri']].plot('label',kind='bar') | _____no_output_____ | Apache-2.0 | annif.ipynb | CSCfi/annif-utils |
4. Perform subject indexing for your own text using project 'yso-omikuji-parabel-en' This can be accomplished using swagger API POST call>**curl command**: curl -X POST --header 'Content-Type: application/x-www-form-urlencoded' --header 'Accept: application/json' -d 'text=frequently occurring or otherwise salient terms in the document are matched with terms in the vocabulary&limit=10' 'https://annif.rahtiapp.fi/v1/projects/yso-omikuji-parabel-en/suggest'Using python approach: | projectid='yso-omikuji-parabel-en'
text='frequently occurring or otherwise salient terms in the document are matched with terms in the vocabulary'
url= base_url+ '/' + projectid +'/suggest'
data = {'text': text}
headers = {'Content-Type': 'application/x-www-form-urlencoded','Accept': 'application/json'}
response = requests.post(url, headers=headers, data=data)
d=response.json() # print(response.text)
display(json_normalize(d['results']))
data=json_normalize(d['results'])
data.loc[:,['label','score','uri']].plot('label',kind='bar') | _____no_output_____ | Apache-2.0 | annif.ipynb | CSCfi/annif-utils |
5. Perform subject indexing for your own text using project 'yso-omikuji-bonsai-en' This can be accomplished using swagger API POST call>**curl command**:curl -X POST --header 'Content-Type: application/x-www-form-urlencoded' --header 'Accept: application/json' -d 'text=frequently occurring or otherwise salient terms in the document are matched with terms in the vocabulary&limit=10' 'https://annif.rahtiapp.fi/v1/projects/yyso-omikuji-bonsai-en/suggest'Using python approach: | projectid='yso-omikuji-bonsai-en'
text='frequently occurring or otherwise salient terms in the document are matched with terms in the vocabulary'
url= base_url+ '/' + projectid +'/suggest'
data = {'text': text}
headers = {'Content-Type': 'application/x-www-form-urlencoded','Accept': 'application/json'}
response = requests.post(url, headers=headers, data=data)
d=response.json() # print(response.text)
display(json_normalize(d['results']))
data=json_normalize(d['results'])
data.loc[:,['label','score','uri']].plot('label',kind='bar') | _____no_output_____ | Apache-2.0 | annif.ipynb | CSCfi/annif-utils |
6. Perform subject indexing for your own text using project 'yso-nn-ensemble-en' This can be accomplished using swagger API POST call>**curl command**: curl -X POST --header 'Content-Type: application/x-www-form-urlencoded' --header 'Accept: application/json' -d 'text=frequently occurring or otherwise salient terms in the document are matched with terms in the vocabulary&limit=10' 'https://annif.rahtiapp.fi/v1/projects/yso-nn-ensemble-en/suggest'Using python approach: | projectid='yso-nn-ensemble-en'
text='frequently occurring or otherwise salient terms in the document are matched with terms in the vocabulary'
url= base_url+ '/' + projectid +'/suggest'
data = {'text': text}
headers = { 'Content-Type': 'application/x-www-form-urlencoded','Accept': 'application/json'}
response = requests.post(url, headers=headers, data=data)
d=response.json()
display(json_normalize(d['results']))
data=json_normalize(d['results'])
data.loc[:,['label','score','uri']].plot('label',kind='bar') | _____no_output_____ | Apache-2.0 | annif.ipynb | CSCfi/annif-utils |
Several exercises: Jupyter Notebook, iPython and ipyparallel, and HPC MPICHThe content of this notebook is borrowed extensively from Daan Van Hauwermeiren, from his tutotial of ipyparallel, stored on github https://github.com/DaanVanHauwermeiren/ipyparallel-tutorial/blob/master/02-ipyparallel-tutorial-direct-interface.ipynb. Many adaptions have been made to accommodate this demonstration and this HPC MPI environment.Prior to running these steps in this notebook, the following details mustbe completed outside the context of Jupyter, and generally will need to be facilitated by a systems administator with appropriate rights and knowledge of HPC and MPI. 1) HPC and MPICH have been configured, with compute nodes running 2) Related to HPC and MPICH, an NFS/NIS environment exists to facilitate the 'scientist' user environment across all of the computing resources in the cluster 3) The ipyparallel client/engine environment must be configured and started that supports MPI/MPICH 4) Ensure that the "IPython Cluster" called "mpi" in the JupyterHub is runningOnce the above details have been acomplished, import the IPython ipyparallel module and create a Client instance | # import the IPython ipyparallel module and create a Client instance
# In this demonstration, an MPI-oriented client is created, referenced by the 'mpi' profile
# There are 4 mpi engines that have been configured and running on 4 separate HPC compute nodes
import ipyparallel as ipp
rc = ipp.Client(profile='mpi')
# Show that there are engines running, responding
rc.ids
# Create an ipyparallel object, constructed via list-access to the client:
vobject = rc[:] | _____no_output_____ | CC0-1.0 | mpi.ipynb | craiggardner/jupyter_notebooks |
Pythonβs builtin map() functions allows a function to be applied to a sequence element-by-element. This type of code is typically trivial to parallelize. In fact, since IPythonβs interface is all about functions anyway, you can just use the builtin map() with a RemoteFunction, or a vobjectβs map() method.do an arbitrary serial computation using just the power of the HPC head node, ... and show how long it takes to compute | %%time
serial_result = list(map(lambda x:x**2**2, range(30))) | CPU times: user 18 Β΅s, sys: 3 Β΅s, total: 21 Β΅s
Wall time: 25.5 Β΅s
| CC0-1.0 | mpi.ipynb | craiggardner/jupyter_notebooks |
Now do the same computation using the MPI compute nodes HPC cluster, ... and show how long it takes | %%time
parallel_result = vobject.map_sync(lambda x:x**2**2, range(30))
serial_result==parallel_result | _____no_output_____ | CC0-1.0 | mpi.ipynb | craiggardner/jupyter_notebooks |
Remote function decoratorsRemote functions are just like normal functions, but when they are called, they execute on one or more engines, rather than locally. Here we will demonstrate the @parallel function decorator, which creates parallel functions that break up an element-wise operations and distribute them to remote workers. It also reconstructs the result from each worker as the result is returned. | # First, we'll enable blocking, which will be explored more throughly later.
# In short, blocking will ensure that each task won't proceed until all the remotely distributed work is complete.
@vobject.remote(block=True)
# Define a function called "getpid" that ... well, you can see the description
def getpid():
'''
import library os and return the process number (pid) corresponding with
the execution
'''
import os
return os.getpid()
# Using our newly defined function, show the process id of the engine running on each compute node
getpid() | _____no_output_____ | CC0-1.0 | mpi.ipynb | craiggardner/jupyter_notebooks |
We'll use numpy to create some complicated (random) arrays, then use those arrays for some big computations that should benefit by some distributed HPC compute resources. | import numpy as np
A = np.random.random((64,48))
# Create a little function that can do the calculations as a distribution among multiple, parallel compute nodes
@vobject.parallel(block=True)
def pmul(A,B):
return A*B | _____no_output_____ | CC0-1.0 | mpi.ipynb | craiggardner/jupyter_notebooks |
We want to be able to compare the amount of time it takes to do the calulation locally on the HPC headand the amount of time it takes to do the calculation among the distributed compute nodesFirst, do the calculation locally, then do it remotely | %%time
C_local = A*A
%%time
C_remote = pmul(A,A)
(C_local == C_remote).all() | _____no_output_____ | CC0-1.0 | mpi.ipynb | craiggardner/jupyter_notebooks |
Create a simple, new function that can be called locally but that will execute remotely, in parallel.It's just a simple instruction that will "echo" the output of what is run on the remote worker | @vobject.parallel(block=True)
def echo(x):
return str(x)
echo(range(5))
echo.map(range(5)) | _____no_output_____ | CC0-1.0 | mpi.ipynb | craiggardner/jupyter_notebooks |
Blocking executionIn blocking mode, the iPython ipyparallel object (called vobject in these examples; defined at the beginning of this notebook) submits the command to the controller, which places the command in the enginesβ queues for execution. The apply() call then blocks until the engines are done executing the command. | # Show function names (on the remote worker) that beging with the string "apply"
[x for x in dir(vobject) if x.startswith('apply')]
vobject.block = True
vobject['a'] = 5
vobject['b'] = 10
vobject.apply(lambda x: a+b+x, 27)
vobject.block = False
vobject.apply_sync(lambda x: a+b+x, 27) | _____no_output_____ | CC0-1.0 | mpi.ipynb | craiggardner/jupyter_notebooks |
Python commands can be executed as strings on specific engines by using a vobjectβs execute method: | rc[::2].execute('c=a+b')
rc[1::2].execute('c=a-b')
vobject['c'] # shorthand for vobject.pull('c', block=True) | _____no_output_____ | CC0-1.0 | mpi.ipynb | craiggardner/jupyter_notebooks |
Non-blocking executionIn non-blocking mode, apply() submits the command to be executed and then returns a AsyncResult object immediately. The AsyncResult object gives you a way of getting a result at a later time through its get() method.More info on the AsyncResult object: http://ipyparallel.readthedocs.io/en/6.0.2/asyncresult.htmlparallel-asyncresultThis allows you to quickly submit long running commands without blocking your local Python/IPython session: | # define our function
def wait(t):
import time
tic = time.time()
time.sleep(t)
return time.time()-tic
# In non-blocking mode
ar = vobject.apply_async(wait, 3)
# Now block for the result, and the output won't disply until after 3 seconds
ar.get()
# Again in non-blocking mode, with longer wait (10 seconds)
ar = vobject.apply_async(wait, 10)
# Poll to see if the result is ready
# If you run this fast enough following the previous step, the output will be "False"
# But if you wait for 10 seconds, before executing this step, the output will be "True"
ar.ready()
# ask for the result, but wait a maximum of 1 second:
ar.get(1) | _____no_output_____ | CC0-1.0 | mpi.ipynb | craiggardner/jupyter_notebooks |
Often, it is desirable to wait until a set of AsyncResult objects are done. For this, there is the method wait(). This method takes a tuple of AsyncResult objects (or msg_ids or indices to the clientβs History), and blocks until all of the associated results are ready.In proper Jupyter Notebook fashion, the step progress indicator will show as a '*' character until the instruction is completed. Output will not be displayed until the instruction is completed. | vobject.block=False
# A trivial list of AsyncResults objects
pr_list = [vobject.apply_async(wait, 3) for i in range(10)]
# Wait until all of the clients have completed the instruction
vobject.wait(pr_list)
# Then, their results are ready using get() or the `.r` attribute
pr_list[0].get() | _____no_output_____ | CC0-1.0 | mpi.ipynb | craiggardner/jupyter_notebooks |
Scatter and gatherSometimes it is useful to partition a sequence and push the partitions to different engines. In MPI language, this is know as scatter/gather and we follow that terminology. However, it is important to remember that in IPythonβs Client class, scatter() is from the interactive IPython session to the engines and gather() is from the engines back to the interactive IPython session. For scatter/gather operations between engines, MPI, pyzmq, or some other direct interconnect should be used. | vobject.scatter('a',range(16))
vobject['a']
vobject.gather('a') | _____no_output_____ | CC0-1.0 | mpi.ipynb | craiggardner/jupyter_notebooks |
parallel list comprehensionsIn many cases list comprehensions are nicer than using the map function. While we donβt have fully parallel list comprehensions, it is simple to get the basic effect using scatter() and gather():The %px magic executes a single Python command on the engines specified by the targets attribute of the view instance | vobject.scatter('x', range(64))
#Parallel execution on engines: [0, 1, 2, 3]
%px y = [i**10 for i in x]
y = vobject.gather('y')
print(y.get()[-10:]) | [210832519264920576, 253295162119140625, 303305489096114176, 362033331456891249, 430804206899405824, 511116753300641401, 604661760000000000, 713342911662882601, 839299365868340224, 984930291881790849]
| CC0-1.0 | mpi.ipynb | craiggardner/jupyter_notebooks |
example: monte carlo approximation of piA simple toy problem to get a handle on multiple engines is a Monte Carlo approximation of Ο.Letβs say we have a dartboard with a round target inscribed on a square board. If you threw darts randomly, and they land evenly distributed on the square board, how many darts would you expect to hit the target? | from random import random
from math import pi
vobject['random'] = random
def mcpi(nsamples):
s = 0
for i in range(nsamples):
x = random()
y = random()
if x*x + y*y <= 1:
s+=1
return 4.*s/nsamples
def multi_mcpi(view, nsamples):
p = len(view.targets)
if nsamples % p:
# ensure even divisibility
nsamples += p - (nsamples%p)
subsamples = nsamples//p
ar = view.apply(mcpi, subsamples)
return sum(ar)/p
def check_pi(tol=1e-5, step=10, verbose=False):
guess = 0
spi = pi
steps = 0
while abs(spi-guess)/spi > tol:
for i in range(step):
x = random()
y = random()
if x*x+y*y <= 1:
guess += 4.
steps += step
spi = pi*steps
if verbose:
print(spi, guess, abs(spi-guess)/spi)
return steps, guess/steps
%%time
mcpi(int(1e9))
# 1e7 means 10 to the 7th power. "e" stands for "expontent"
%%time
multi_mcpi(vobject, int(1e9))
check_pi() | _____no_output_____ | CC0-1.0 | mpi.ipynb | craiggardner/jupyter_notebooks |
def what_is_installed():
import pycaret
from pycaret import show_versions
show_versions()
try:
what_is_installed()
except:
!pip install pycaret-ts-alpha
what_is_installed()
import numpy as np
import pandas as pd
from pycaret.datasets import get_data
from pycaret.time_series import TSForecastingExperiment
#### Exogenous variables ----
data = pd.DataFrame({'a': np.random.randn(200), 'b': np.random.randn(200), 'c': np.random.randn(200)})
#### Produce dependent variable based on exogenous variables ----
# NOTE: Only dependent on 'a' and 'c' but not 'b'
data['y'] = data['a'].shift(4) + data['c'].shift(8)
data.dropna(inplace=True)
data.shape
#### Create Time Series Forecasting Experiment ----
exp = TSForecastingExperiment()
global_plot_settings = {"renderer": "colab"}
exp.setup(data=data, target="y", seasonal_period=1, fh=8, fig_kwargs=global_plot_settings, session_id=42)
exp.plot_model(plot="acf") | _____no_output_____ | MIT | time_series/pycaret/pycaret_ts_ccf.ipynb | ngupta23/medium_articles |
|
**Not much to go by in terms of forecasting y just by itself (without exogenous variables)** | exp.plot_model(plot="ccf") | _____no_output_____ | MIT | time_series/pycaret/pycaret_ts_ccf.ipynb | ngupta23/medium_articles |
Q-Learning> Off-Policy Temporal Difference Learning. | import gym
import numpy as np
import matplotlib.pyplot as plt
from IPython.display import clear_output
from time import sleep
env_name = "Taxi-v3"
epsilon = 1
decay_rate = 0.001
min_epsilon = 0.01
max_episodes = 2500
print_interval = 100
test_episodes = 3
lr = 0.4
gamma = 0.99
env = gym.make(env_name)
env = gym.wrappers.Monitor(env, "./vid", force=True)
n_states = env.observation_space.n
n_actions = env.action_space.n
print(f"Number of states: {n_states}\n"
f"Number of actions: {n_actions}")
q_table = np.zeros((n_states, n_actions))
def choose_action(state):
global q_table
if epsilon > np.random.uniform():
action = env.action_space.sample()
else:
action = np.argmax(q_table[state, :])
return action
def update_table(state, action, reward, done, next_state):
global q_table
q_table[state, action] += lr * (reward + gamma * np.max(q_table[next_state, :]) * (1 - done) - q_table[state, action]) | _____no_output_____ | MIT | Q_Learning/Taxi_env.ipynb | alirezakazemipour/Q-Table-Numpy |
Pseudocode> | running_reward = []
for episode in range(1, 1 + max_episodes):
state = env.reset()
done = False
episode_reward = 0
while not done:
action = choose_action(state)
next_state, reward, done, _ = env.step(action)
update_table(state, action, reward, done, next_state)
episode_reward += reward
if done:
break
state = next_state
epsilon = epsilon - decay_rate if epsilon - decay_rate > min_epsilon else min_epsilon
if episode == 1:
running_reward.append(episode_reward)
else:
running_reward.append(0.99 * running_reward[-1] + 0.01 * episode_reward)
if episode % print_interval == 0:
print(f"Ep:{episode}| "
f"Ep_reward:{episode_reward}| "
f"Running_reward:{running_reward[-1]:.3f}| "
f"Epsilon:{epsilon:.3f}| ")
plt.figure()
plt.style.use("ggplot")
plt.plot(np.arange(max_episodes), running_reward)
plt.title("Running_reward")
for episode in range(1, 1 + test_episodes):
state = env.reset()
done = False
episode_reward = 0
while not done:
action = choose_action(state)
next_state, reward, done, _ = env.step(action)
env.render()
clear_output(wait=True)
sleep(0.2)
episode_reward += reward
if done:
break
state = next_state
print(f"Ep:{episode}| "
f"Ep_reward:{episode_reward}| ")
env.close() | Ep:3| Ep_reward:4|
| MIT | Q_Learning/Taxi_env.ipynb | alirezakazemipour/Q-Table-Numpy |
Stationary freatic flow between two water courses above a semi-pervious layer (Wesseling) Constant precipitation N on a strip of land between two parallel water courses with water level hs causes a rise h(x) of the groundwater level that induces in a groundwater flow towards the water courses. The phreatic groundwater is separated from an aquifer below by a semi-pervious layer. The interaction between phreatic groundwater system and the aquifer is determined by the resistance c of the semi-pervious layer and the groundwater head H in the aquifer below. A solution was published by Wesseling & Wesseling (1984). A practical dicussion is provided by Van Drecht (1997, in Dutch) | import numpy as np
%matplotlib inline
import matplotlib.pyplot as plt
plt.style.use('seaborn-whitegrid')
def hx(x,L,hs,p,H,c,T):
"""Return phreatic groundwater level h(x) between two water courses
Parameters:
x : numpy array
Distance from centre of the water course (m)
L : float
Distance between the centre of both water courses (m)
hs : float
water level in the water courses (m)
p : float
precipitation (m/day)
H : float
groundwater head in the deep aquifer (m)
c : float
resistance of semi-pervious layer (day)
T : float
transmissivity of the deep aquifer (m2/day)
Returns
-------
numpy array
"""
labda = np.sqrt(T*c)
alpha = L / (2*labda)
hx = hs + (H - hs + p*c)* (np.tanh(alpha) * np.sinh(x/labda) - np.cosh(x/labda) + 1)
return hx
def qd(L,hs,p,H,c,T):
"""Return groundwater discharge to water courses"""
labda = np.sqrt(T*c)
alpha = L / (2*labda)
qd = ((H-hs)/c+p)*np.tanh(alpha)/alpha
return qd
def qs(L,hs,p,H,c,T):
"""Return groundwater seepage"""
#labda = np.sqrt(T*c)
#alpha = L / (2*labda)
#qd = -p + ((H-hs)/c+p)*np.tanh(alpha)/alpha
qs = -p + qd(L,hs,p,H,c,T)
return qs
L = 400.
hs = 0.0
p = 300./365./1000.
H = 0.0
c = 100000.
T = 15.
x = np.linspace(1,L/2,25)
# h0
p = 300./365./1000.
H = 0.0
c = 10.
h0 = hx(x,L,hs,p,H,c,T)
qd0 = qd(L,hs,p,H,c,T)
qs0 = qs(L,hs,p,H,c,T)
# h1
p = 300./365./1000.
H = -0.025
c = 10.
h1 = hx(x,L,hs,p,H,c,T)
qd1 = qd(L,hs,p,H,c,T)
qs1 = qs(L,hs,p,H,c,T)
# h2
p = 300./365./1000.
H = 0.025
c = 10.
h2 = hx(x,L,hs,p,H,c,T)
qd2 = qd(L,hs,p,H,c,T)
qs2 = qs(L,hs,p,H,c,T)
# h3
p = 7./1000.
H = 0.0
c = 10.
h3 = hx(x,L,hs,p,H,c,T)
qd3 = qd(L,hs,p,H,c,T)
qs3 = qs(L,hs,p,H,c,T)
# h4
p = 300./365./1000.
H = 0.20
c = 10.
h4 = hx(x,L,hs,p,H,c,T)
qd4 = qd(L,hs,p,H,c,T)
qs4 = qs(L,hs,p,H,c,T)
# h5
p = 7./1000.
H = 0.20
c = 10.
h5 = hx(x,L,hs,p,H,c,T)
qd5 = qd(L,hs,p,H,c,T)
qs5 = qs(L,hs,p,H,c,T)
plt.rcParams['figure.figsize'] = [10, 7]
fig, ax = plt.subplots()
#4c4c4c gray
#f65b00 orange
#fabb01 yellow
#0000FF dark blue
#0080FF light blue
ax.plot(x, h0, '-', color='#0080FF', label='dH = 0, N = 0.8 mm/d');
ax.plot(x, h3, '-', color='#0000FF', label='dH = 0cm, N = 7 mm/d');
ax.plot(x, h1, '--', color='#0080FF', label='dH = -2.5cm, N = 0.8 mm/d');
ax.plot(x, h2, '--', color='#0080FF', label='dH = +2.5cm, N = 0.8 mm/d');
ax.plot(x, h4, '-', color='#fabb01', label='dH = +20cm, N = 0.8 mm/d');
ax.plot(x, h5, '-', color='#f65b00', label='dH = +20cm, N = 7 mm/d');
ax.set_ylim(-0.05, 0.3)
ax.set_xlim(0, L/2)
ax.set_xlabel('afstand tot de beek (m)',fontsize=11)
ax.set_ylabel('opbolling van het grondwater (m)', fontsize = 11)
#ax.set_title('')
#lg = ax.legend(loc="upper right")
ax.grid(True)
#legframe = plt.legend.get_frame()
#lg.get_frame().set_alpha(0.5)
#lg.get_frame().set_facecolor('white')
ax.legend(loc="upper right", fancybox=True, frameon=True, facecolor='white', framealpha=1, )
plt.show()
fig.savefig('wesseling_hx.png')
# slootafvoer
for q in [0.8/1000,7/1000,qd0,qd3,qd4,qd5]:
m3dagkm = q*L*1000
lsdkm = m3dagkm*1/86.4
print(m3dagkm,lsdkm)
# kwelflux
for q in [0.8/1000,7/1000,qs0,qs3,qs4,qs5]:
m3dagkm = q*L*1000
lsdkm = m3dagkm*1/86.4
print(m3dagkm,lsdkm) | 320.0 3.7037037037037033
2800.0000000000005 32.40740740740741
-308.63433088123435 -3.5721566074216935
-2628.53571800518 -30.422867106541435
181.26361767539473 2.097958537909661
-2138.637769448551 -24.752751961210077
| MIT | Watercourse/Stationary freatic flow between two water courses above a semi-pervious layer (Wesseling).ipynb | tdmeij/GWF |
Text Classification with PySpark - Multiclass Text ClassificationTask- Predict the subject category given a course title or text | import pyspark
from pyspark import SparkContext
sc = SparkContext(master='local[2]')
# lunch UI
sc
# create spark seassion
from pyspark.sql import SparkSession
spark = SparkSession.builder.appName("Text Classifier").getOrCreate()
# read the dataset and load
df = spark.read.csv('udemy.csv',header=True, inferSchema=True)
df.show(5)
df.columns
df = df.select('course_title','subject')
df.show(5)
df.groupby('subject').count().sort("count",ascending=False).show()
# getting values count using pandas
# df.toPandas()['subject'].value_counts()
# check for missing values
df.toPandas()['subject'].isnull().sum()
# drop missing values
df = df.dropna(subset= ['subject'])
# check for missing values
df.toPandas()['subject'].isnull().sum()
df.show(5) | +--------------------+----------------+
| course_title| subject|
+--------------------+----------------+
|Ultimate Investme...|Business Finance|
|Complete GST Cour...|Business Finance|
|Financial Modelin...|Business Finance|
|Beginner to Pro -...|Business Finance|
|How To Maximize Y...|Business Finance|
+--------------------+----------------+
only showing top 5 rows
| Apache-2.0 | getStarted/TextClassification/textClassification.ipynb | iamhimanshu0/Spark |
Feature Extractionbuild features + count vectorizer+ tfIDF+ wordEmbeddings+ hashingTF+ etc...We have 2 things in Pipeline stages- Transformer- Estimator**Transformer** (Data to Data)Function that takes data and fit, transform them into augmented data or featuresi.e Extractors, Vectorizer, Scalers (Tokenizer, StopwordRemover, CountVectorizer, IDF)**Estimator** (Data to model)Function that takes data as input and fit the data and produces a model we can use to predicti.e LogisticRegression | from pyspark.ml.feature import Tokenizer, StopWordsRemover, CountVectorizer, IDF, StringIndexer
# dir(pyspark.ml.feature)
# Stages for the pipeline
tokenizer = Tokenizer(inputCol='course_title', outputCol='mytokens')
stopwordRemover = StopWordsRemover(inputCol='mytokens',outputCol='filtered_tokens')
vectorizer = CountVectorizer(inputCol='filtered_tokens',outputCol='rawFeatures')
idf = IDF(inputCol='rawFeatures', outputCol='vectorizedFeatures')
# work on taget variable (subject)
# label encoding/indexing
labelEncoder = StringIndexer(inputCol='subject',outputCol='label').fit(df)
labelEncoder.transform(df).show(5)
# labelEncoder.labels
# making dict to labels
label_dict = {
'Web Development':0.0,
'Business Finance':1.0,
'Musical Instruments':2.0,
'Graphic Design':3.0
}
df = labelEncoder.transform(df)
df.show(5)
# split dataset
(train_df, test_df) = df.randomSplit((0.7,0.3),seed=42)
train_df.show(2)
# machine learning model (Estimator) (data to model)
from pyspark.ml.classification import LogisticRegression
lr = LogisticRegression(featuresCol='vectorizedFeatures',
labelCol = 'label'
) | _____no_output_____ | Apache-2.0 | getStarted/TextClassification/textClassification.ipynb | iamhimanshu0/Spark |
Building the pipeline | from pyspark.ml import Pipeline
pipeline = Pipeline(
stages=[tokenizer, stopwordRemover, vectorizer, idf, lr]
)
pipeline.stages
# model building
lr_model = pipeline.fit(train_df)
lr_model
# get predicction on test data
predictions = lr_model.transform(test_df)
# predictions.show()
predictions.columns
predictions.select('rawPrediction', 'probability','subject','label','prediction').show(10) | +--------------------+--------------------+-------------------+-----+----------+
| rawPrediction| probability| subject|label|prediction|
+--------------------+--------------------+-------------------+-----+----------+
|[8.30964874634511...|[0.87877993991729...|Musical Instruments| 2.0| 0.0|
|[-1.3744065857781...|[1.90975343878318...|Musical Instruments| 2.0| 2.0|
|[0.60822716351824...|[3.28451283099288...|Musical Instruments| 2.0| 2.0|
|[-1.0584564885297...|[3.70732079181542...| Business Finance| 1.0| 1.0|
|[24.6296077836821...|[0.99999999906211...| Web Development| 0.0| 0.0|
|[22.0136686708729...|[0.99999999049941...| Web Development| 0.0| 0.0|
|[19.9225858177008...|[0.99999995276066...| Web Development| 0.0| 0.0|
|[-5.7386799100009...|[5.78822181193782...|Musical Instruments| 2.0| 2.0|
|[-19.060576929776...|[1.71813778453453...| Graphic Design| 3.0| 3.0|
|[-2.4736166619785...|[1.84538870784594...|Musical Instruments| 2.0| 2.0|
+--------------------+--------------------+-------------------+-----+----------+
only showing top 10 rows
| Apache-2.0 | getStarted/TextClassification/textClassification.ipynb | iamhimanshu0/Spark |
model evaluation+ Accuracy+ Precision+ F1Score+ etc | from pyspark.ml.evaluation import MulticlassClassificationEvaluator
evaluator = MulticlassClassificationEvaluator(predictionCol='prediction',labelCol='label')
accuracy = evaluator.evaluate(predictions)
accuracy*100
"""
Method 2:
precision, f1score classification report
"""
from pyspark.mllib.evaluation import MulticlassMetrics
lr_metric = MulticlassMetrics(predictions['label','prediction'].rdd)
print("Accuracy ", lr_metric.accuracy)
print("precision ", lr_metric.precision(1.0))
print("f1Score ", lr_metric.fMeasure(1.0))
print("recall ", lr_metric.recall(1.0)) | Accuracy 0.9182509505703422
precision 0.9544159544159544
f1Score 0.9178082191780822
recall 0.8839050131926122
| Apache-2.0 | getStarted/TextClassification/textClassification.ipynb | iamhimanshu0/Spark |
Confusion matrix- convert to pandas- sklearn | y_true = predictions.select('label')
y_true = y_true.toPandas()
y_predict = predictions.select('prediction')
y_predict = y_predict.toPandas()
from sklearn.metrics import confusion_matrix, classification_report
cm = confusion_matrix(y_true, y_predict)
cm | _____no_output_____ | Apache-2.0 | getStarted/TextClassification/textClassification.ipynb | iamhimanshu0/Spark |
making prediction on one sample+ sample as df+ apply pipeline | from pyspark.sql.types import StringType
exl = spark.createDataFrame([
("Building Machine Learning Apps with Python and PySpark", StringType())
],
#column name
['course_title']
)
exl.show()
# show fill
exl.show(truncate=False)
# making prediction
prediction_ex1 = lr_model.transform(exl)
prediction_ex1.show(truncate=True)
prediction_ex1.columns
prediction_ex1.select('course_title','rawPrediction','probability','prediction').show()
label_dict
# save and load the model
modelPath = "models/pyspark_lr_model"
lr_model.write().save(modelPath)
# loading pickled model
from pyspark.ml.pipeline import PipelineModel
presistedModel =PipelineModel.load(modelPath)
# laodModel
# making prediction
loadModel = presistedModel.transform(exl)
loadModel.show(truncate=True)
loadModel.select('course_title','rawPrediction','probability','prediction').show() | +--------------------+--------------------+--------------------+----------+
| course_title| rawPrediction| probability|prediction|
+--------------------+--------------------+--------------------+----------+
|Building Machine ...|[14.7174498131555...|[0.99999814636182...| 0.0|
+--------------------+--------------------+--------------------+----------+
| Apache-2.0 | getStarted/TextClassification/textClassification.ipynb | iamhimanshu0/Spark |
______Copyright by Pierian Data Inc.For more information, visit us at www.pieriandata.com Text Methods A normal Python string has a variety of method calls available: | mystring = 'hello'
mystring.capitalize()
mystring.isdigit()
help(str) | Help on class str in module builtins:
class str(object)
| str(object='') -> str
| str(bytes_or_buffer[, encoding[, errors]]) -> str
|
| Create a new string object from the given object. If encoding or
| errors is specified, then the object must expose a data buffer
| that will be decoded using the given encoding and error handler.
| Otherwise, returns the result of object.__str__() (if defined)
| or repr(object).
| encoding defaults to sys.getdefaultencoding().
| errors defaults to 'strict'.
|
| Methods defined here:
|
| __add__(self, value, /)
| Return self+value.
|
| __contains__(self, key, /)
| Return key in self.
|
| __eq__(self, value, /)
| Return self==value.
|
| __format__(self, format_spec, /)
| Return a formatted version of the string as described by format_spec.
|
| __ge__(self, value, /)
| Return self>=value.
|
| __getattribute__(self, name, /)
| Return getattr(self, name).
|
| __getitem__(self, key, /)
| Return self[key].
|
| __getnewargs__(...)
|
| __gt__(self, value, /)
| Return self>value.
|
| __hash__(self, /)
| Return hash(self).
|
| __iter__(self, /)
| Implement iter(self).
|
| __le__(self, value, /)
| Return self<=value.
|
| __len__(self, /)
| Return len(self).
|
| __lt__(self, value, /)
| Return self<value.
|
| __mod__(self, value, /)
| Return self%value.
|
| __mul__(self, value, /)
| Return self*value.
|
| __ne__(self, value, /)
| Return self!=value.
|
| __repr__(self, /)
| Return repr(self).
|
| __rmod__(self, value, /)
| Return value%self.
|
| __rmul__(self, value, /)
| Return value*self.
|
| __sizeof__(self, /)
| Return the size of the string in memory, in bytes.
|
| __str__(self, /)
| Return str(self).
|
| capitalize(self, /)
| Return a capitalized version of the string.
|
| More specifically, make the first character have upper case and the rest lower
| case.
|
| casefold(self, /)
| Return a version of the string suitable for caseless comparisons.
|
| center(self, width, fillchar=' ', /)
| Return a centered string of length width.
|
| Padding is done using the specified fill character (default is a space).
|
| count(...)
| S.count(sub[, start[, end]]) -> int
|
| Return the number of non-overlapping occurrences of substring sub in
| string S[start:end]. Optional arguments start and end are
| interpreted as in slice notation.
|
| encode(self, /, encoding='utf-8', errors='strict')
| Encode the string using the codec registered for encoding.
|
| encoding
| The encoding in which to encode the string.
| errors
| The error handling scheme to use for encoding errors.
| The default is 'strict' meaning that encoding errors raise a
| UnicodeEncodeError. Other possible values are 'ignore', 'replace' and
| 'xmlcharrefreplace' as well as any other name registered with
| codecs.register_error that can handle UnicodeEncodeErrors.
|
| endswith(...)
| S.endswith(suffix[, start[, end]]) -> bool
|
| Return True if S ends with the specified suffix, False otherwise.
| With optional start, test S beginning at that position.
| With optional end, stop comparing S at that position.
| suffix can also be a tuple of strings to try.
|
| expandtabs(self, /, tabsize=8)
| Return a copy where all tab characters are expanded using spaces.
|
| If tabsize is not given, a tab size of 8 characters is assumed.
|
| find(...)
| S.find(sub[, start[, end]]) -> int
|
| Return the lowest index in S where substring sub is found,
| such that sub is contained within S[start:end]. Optional
| arguments start and end are interpreted as in slice notation.
|
| Return -1 on failure.
|
| format(...)
| S.format(*args, **kwargs) -> str
|
| Return a formatted version of S, using substitutions from args and kwargs.
| The substitutions are identified by braces ('{' and '}').
|
| format_map(...)
| S.format_map(mapping) -> str
|
| Return a formatted version of S, using substitutions from mapping.
| The substitutions are identified by braces ('{' and '}').
|
| index(...)
| S.index(sub[, start[, end]]) -> int
|
| Return the lowest index in S where substring sub is found,
| such that sub is contained within S[start:end]. Optional
| arguments start and end are interpreted as in slice notation.
|
| Raises ValueError when the substring is not found.
|
| isalnum(self, /)
| Return True if the string is an alpha-numeric string, False otherwise.
|
| A string is alpha-numeric if all characters in the string are alpha-numeric and
| there is at least one character in the string.
|
| isalpha(self, /)
| Return True if the string is an alphabetic string, False otherwise.
|
| A string is alphabetic if all characters in the string are alphabetic and there
| is at least one character in the string.
|
| isascii(self, /)
| Return True if all characters in the string are ASCII, False otherwise.
|
| ASCII characters have code points in the range U+0000-U+007F.
| Empty string is ASCII too.
|
| isdecimal(self, /)
| Return True if the string is a decimal string, False otherwise.
|
| A string is a decimal string if all characters in the string are decimal and
| there is at least one character in the string.
|
| isdigit(self, /)
| Return True if the string is a digit string, False otherwise.
|
| A string is a digit string if all characters in the string are digits and there
| is at least one character in the string.
|
| isidentifier(self, /)
| Return True if the string is a valid Python identifier, False otherwise.
|
| Use keyword.iskeyword() to test for reserved identifiers such as "def" and
| "class".
|
| islower(self, /)
| Return True if the string is a lowercase string, False otherwise.
|
| A string is lowercase if all cased characters in the string are lowercase and
| there is at least one cased character in the string.
|
| isnumeric(self, /)
| Return True if the string is a numeric string, False otherwise.
|
| A string is numeric if all characters in the string are numeric and there is at
| least one character in the string.
|
| isprintable(self, /)
| Return True if the string is printable, False otherwise.
|
| A string is printable if all of its characters are considered printable in
| repr() or if it is empty.
|
| isspace(self, /)
| Return True if the string is a whitespace string, False otherwise.
|
| A string is whitespace if all characters in the string are whitespace and there
| is at least one character in the string.
|
| istitle(self, /)
| Return True if the string is a title-cased string, False otherwise.
|
| In a title-cased string, upper- and title-case characters may only
| follow uncased characters and lowercase characters only cased ones.
|
| isupper(self, /)
| Return True if the string is an uppercase string, False otherwise.
|
| A string is uppercase if all cased characters in the string are uppercase and
| there is at least one cased character in the string.
|
| join(self, iterable, /)
| Concatenate any number of strings.
|
| The string whose method is called is inserted in between each given string.
| The result is returned as a new string.
|
| Example: '.'.join(['ab', 'pq', 'rs']) -> 'ab.pq.rs'
|
| ljust(self, width, fillchar=' ', /)
| Return a left-justified string of length width.
|
| Padding is done using the specified fill character (default is a space).
|
| lower(self, /)
| Return a copy of the string converted to lowercase.
|
| lstrip(self, chars=None, /)
| Return a copy of the string with leading whitespace removed.
|
| If chars is given and not None, remove characters in chars instead.
|
| partition(self, sep, /)
| Partition the string into three parts using the given separator.
|
| This will search for the separator in the string. If the separator is found,
| returns a 3-tuple containing the part before the separator, the separator
| itself, and the part after it.
|
| If the separator is not found, returns a 3-tuple containing the original string
| and two empty strings.
|
| replace(self, old, new, count=-1, /)
| Return a copy with all occurrences of substring old replaced by new.
|
| count
| Maximum number of occurrences to replace.
| -1 (the default value) means replace all occurrences.
|
| If the optional argument count is given, only the first count occurrences are
| replaced.
|
| rfind(...)
| S.rfind(sub[, start[, end]]) -> int
|
| Return the highest index in S where substring sub is found,
| such that sub is contained within S[start:end]. Optional
| arguments start and end are interpreted as in slice notation.
|
| Return -1 on failure.
|
| rindex(...)
| S.rindex(sub[, start[, end]]) -> int
|
| Return the highest index in S where substring sub is found,
| such that sub is contained within S[start:end]. Optional
| arguments start and end are interpreted as in slice notation.
|
| Raises ValueError when the substring is not found.
|
| rjust(self, width, fillchar=' ', /)
| Return a right-justified string of length width.
|
| Padding is done using the specified fill character (default is a space).
|
| rpartition(self, sep, /)
| Partition the string into three parts using the given separator.
|
| This will search for the separator in the string, starting at the end. If
| the separator is found, returns a 3-tuple containing the part before the
| separator, the separator itself, and the part after it.
|
| If the separator is not found, returns a 3-tuple containing two empty strings
| and the original string.
|
| rsplit(self, /, sep=None, maxsplit=-1)
| Return a list of the words in the string, using sep as the delimiter string.
|
| sep
| The delimiter according which to split the string.
| None (the default value) means split according to any whitespace,
| and discard empty strings from the result.
| maxsplit
| Maximum number of splits to do.
| -1 (the default value) means no limit.
|
| Splits are done starting at the end of the string and working to the front.
|
| rstrip(self, chars=None, /)
| Return a copy of the string with trailing whitespace removed.
|
| If chars is given and not None, remove characters in chars instead.
|
| split(self, /, sep=None, maxsplit=-1)
| Return a list of the words in the string, using sep as the delimiter string.
|
| sep
| The delimiter according which to split the string.
| None (the default value) means split according to any whitespace,
| and discard empty strings from the result.
| maxsplit
| Maximum number of splits to do.
| -1 (the default value) means no limit.
|
| splitlines(self, /, keepends=False)
| Return a list of the lines in the string, breaking at line boundaries.
|
| Line breaks are not included in the resulting list unless keepends is given and
| true.
|
| startswith(...)
| S.startswith(prefix[, start[, end]]) -> bool
|
| Return True if S starts with the specified prefix, False otherwise.
| With optional start, test S beginning at that position.
| With optional end, stop comparing S at that position.
| prefix can also be a tuple of strings to try.
|
| strip(self, chars=None, /)
| Return a copy of the string with leading and trailing whitespace removed.
|
| If chars is given and not None, remove characters in chars instead.
|
| swapcase(self, /)
| Convert uppercase characters to lowercase and lowercase characters to uppercase.
|
| title(self, /)
| Return a version of the string where each word is titlecased.
|
| More specifically, words start with uppercased characters and all remaining
| cased characters have lower case.
|
| translate(self, table, /)
| Replace each character in the string using the given translation table.
|
| table
| Translation table, which must be a mapping of Unicode ordinals to
| Unicode ordinals, strings, or None.
|
| The table must implement lookup/indexing via __getitem__, for instance a
| dictionary or list. If this operation raises LookupError, the character is
| left untouched. Characters mapped to None are deleted.
|
| upper(self, /)
| Return a copy of the string converted to uppercase.
|
| zfill(self, width, /)
| Pad a numeric string with zeros on the left, to fill a field of the given width.
|
| The string is never truncated.
|
| ----------------------------------------------------------------------
| Static methods defined here:
|
| __new__(*args, **kwargs) from builtins.type
| Create and return a new object. See help(type) for accurate signature.
|
| maketrans(x, y=None, z=None, /)
| Return a translation table usable for str.translate().
|
| If there is only one argument, it must be a dictionary mapping Unicode
| ordinals (integers) or characters to Unicode ordinals, strings or None.
| Character keys will be then converted to ordinals.
| If there are two arguments, they must be strings of equal length, and
| in the resulting dictionary, each character in x will be mapped to the
| character at the same position in y. If there is a third argument, it
| must be a string, whose characters will be mapped to None in the result.
| Apache-2.0 | 07-Text-Methods.ipynb | srijikabanerjee/demo2 |
Pandas and TextPandas can do a lot more than what we show here. Full online documentation on things like advanced string indexing and regular expressions with pandas can be found here: https://pandas.pydata.org/docs/user_guide/text.html Text Methods on Pandas String Column | import pandas as pd
names = pd.Series(['andrew','bobo','claire','david','4'])
names
names.str.capitalize()
names.str.isdigit() | _____no_output_____ | Apache-2.0 | 07-Text-Methods.ipynb | srijikabanerjee/demo2 |
Splitting , Grabbing, and Expanding | tech_finance = ['GOOG,APPL,AMZN','JPM,BAC,GS']
len(tech_finance)
tickers = pd.Series(tech_finance)
tickers
tickers.str.split(',')
tickers.str.split(',').str[0]
tickers.str.split(',',expand=True) | _____no_output_____ | Apache-2.0 | 07-Text-Methods.ipynb | srijikabanerjee/demo2 |
Cleaning or Editing Strings | messy_names = pd.Series(["andrew ","bo;bo"," claire "])
# Notice the "mis-alignment" on the right hand side due to spacing in "andrew " and " claire "
messy_names
messy_names.str.replace(";","")
messy_names.str.strip()
messy_names.str.replace(";","").str.strip()
messy_names.str.replace(";","").str.strip().str.capitalize() | _____no_output_____ | Apache-2.0 | 07-Text-Methods.ipynb | srijikabanerjee/demo2 |
Alternative with Custom apply() call | def cleanup(name):
name = name.replace(";","")
name = name.strip()
name = name.capitalize()
return name
messy_names
messy_names.apply(cleanup) | _____no_output_____ | Apache-2.0 | 07-Text-Methods.ipynb | srijikabanerjee/demo2 |
Which one is more efficient? | import timeit
# code snippet to be executed only once
setup = '''
import pandas as pd
import numpy as np
messy_names = pd.Series(["andrew ","bo;bo"," claire "])
def cleanup(name):
name = name.replace(";","")
name = name.strip()
name = name.capitalize()
return name
'''
# code snippet whose execution time is to be measured
stmt_pandas_str = '''
messy_names.str.replace(";","").str.strip().str.capitalize()
'''
stmt_pandas_apply = '''
messy_names.apply(cleanup)
'''
stmt_pandas_vectorize='''
np.vectorize(cleanup)(messy_names)
'''
timeit.timeit(setup = setup,
stmt = stmt_pandas_str,
number = 10000)
timeit.timeit(setup = setup,
stmt = stmt_pandas_apply,
number = 10000)
timeit.timeit(setup = setup,
stmt = stmt_pandas_vectorize,
number = 10000) | _____no_output_____ | Apache-2.0 | 07-Text-Methods.ipynb | srijikabanerjee/demo2 |
Texas updates their data daily at noon CDT | from selenium import webdriver
import time
import pandas as pd
import pendulum
import re
import yaml
from selenium.webdriver.chrome.options import Options
chrome_options = Options()
#chrome_options.add_argument("--disable-extensions")
#chrome_options.add_argument("--disable-gpu")
#chrome_options.add_argument("--no-sandbox) # linux only
chrome_options.add_argument("--start-maximized")
# chrome_options.add_argument("--headless")
chrome_options.add_argument("user-agent=[Mozilla/5.0 (Macintosh; Intel Mac OS X 10.14; rv:73.0) Gecko/20100101 Firefox/73.0]")
with open('config.yaml', 'r') as f:
config = yaml.safe_load(f.read())
state = 'TX'
scrape_timestamp = pendulum.now().strftime('%Y%m%d%H%M%S')
# MD positive cases by county
url = 'https://www.dshs.state.tx.us/news/updates.shtm#coronavirus'
url = 'https://txdshs.maps.arcgis.com/apps/opsdashboard/index.html#/ed483ecd702b4298ab01e8b9cafc8b83'
def fetch():
driver = webdriver.Chrome('../20190611 - Parts recommendation/chromedriver', options=chrome_options)
driver.get(url)
time.sleep(5)
datatbl = driver.find_element_by_class_name('feature-list')
datatbl.find_elements_by_class_name('external-html')
datatbl = datatbl.text.split('\n')
data = []
for i in range(0, len(datatbl), 2):
data.append([datatbl[i], datatbl[i+1]])
page_source = driver.page_source
driver.close()
return pd.DataFrame(data, columns=['county','positive_cases']), page_source
def save(df, source):
df.to_csv(f"{config['data_folder']}/{state}_county_{scrape_timestamp}.txt", sep='|', index=False)
with open(f"{config['data_source_backup_folder']}/{state}_county_{scrape_timestamp}.html", 'w') as f:
f.write(source)
def run():
df, source = fetch()
save(df, source) | _____no_output_____ | MIT | TX by county.ipynb | kirbs-/covid-19-dataset |
Convert Dataset FormatsThis recipe demonstrates how to use FiftyOne to convert datasets on disk between common formats. Setup If you haven't already, install FiftyOne: | !pip install fiftyone
import fiftyone as fo | _____no_output_____ | Apache-2.0 | docs/source/recipes/convert_datasets.ipynb | pixta-dev/fiftyone |
If the above import fails due to a `cv2` error, it is an issue with OpenCV in Colab environments. [Follow these instructions to resolve it.](https://github.com/voxel51/fiftyone/issues/1494issuecomment-1003148448). This notebook contains bash commands. To run it as a notebook, you must install the [Jupyter bash kernel](https://github.com/takluyver/bash_kernel) via the command below.Alternatively, you can just copy + paste the code blocks into your shell. | pip install bash_kernel
python -m bash_kernel.install | _____no_output_____ | Apache-2.0 | docs/source/recipes/convert_datasets.ipynb | pixta-dev/fiftyone |
In this recipe we'll use the [FiftyOne Dataset Zoo](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/zoo_datasets.html) to download some open source datasets to work with.Specifically, we'll need [TensorFlow](https://www.tensorflow.org/) and [TensorFlow Datasets](https://www.tensorflow.org/datasets) installed to [access the datasets](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/zoo_datasets.htmlcustomizing-your-ml-backend): | pip install tensorflow tensorflow-datasets | _____no_output_____ | Apache-2.0 | docs/source/recipes/convert_datasets.ipynb | pixta-dev/fiftyone |
Download datasets Download the test split of the [CIFAR-10 dataset](https://www.cs.toronto.edu/~kriz/cifar.html) from the [FiftyOne Dataset Zoo](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/zoo_datasets.html) using the command below: | # Download the test split of CIFAR-10
fiftyone zoo datasets download cifar10 --split test | Downloading split 'test' to '~/fiftyone/cifar10/test'
Downloading https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz to ~/fiftyone/cifar10/tmp-download/cifar-10-python.tar.gz
170500096it [00:04, 35887670.65it/s]
Extracting ~/fiftyone/cifar10/tmp-download/cifar-10-python.tar.gz to ~/fiftyone/cifar10/tmp-download
100% |βββ| 10000/10000 [5.2s elapsed, 0s remaining, 1.8K samples/s]
Dataset info written to '~/fiftyone/cifar10/info.json'
| Apache-2.0 | docs/source/recipes/convert_datasets.ipynb | pixta-dev/fiftyone |
Download the validation split of the [KITTI dataset]( http://www.cvlibs.net/datasets/kitti) from the [FiftyOne Dataset Zoo](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/zoo_datasets.html) using the command below: | # Download the validation split of KITTI
fiftyone zoo datasets download kitti --split validation | Split 'validation' already downloaded
| Apache-2.0 | docs/source/recipes/convert_datasets.ipynb | pixta-dev/fiftyone |
The fiftyone convert command The [FiftyOne CLI](https://voxel51.com/docs/fiftyone/cli/index.html) provides a number of utilities for importing and exporting datasets in a variety of common (or custom) formats.Specifically, the `fiftyone convert` command provides a convenient way to convert datasets on disk between formats by specifying the [fiftyone.types.Dataset](https://voxel51.com/docs/fiftyone/api/fiftyone.types.htmlfiftyone.types.dataset_types.Dataset) type of the input and desired output.FiftyOne provides a collection of [builtin types](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlsupported-formats) that you can use to read/write datasets in common formats out-of-the-box: | Dataset format | Import Supported? | Export Supported? | Conversion Supported? || ---------------------------------------------------------------------------------------------------------------------------------------------------- | ----------------- | ----------------- | --------------------- || [ImageDirectory](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlimagedirectory) | β | β | β || [VideoDirectory](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlvideodirectory) | β | β | β || [FiftyOneImageClassificationDataset](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlfiftyoneimageclassificationdataset) | β | β | β || [ImageClassificationDirectoryTree](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlimageclassificationdirectorytree) | β | β | β || [TFImageClassificationDataset](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmltfimageclassificationdataset) | β | β | β || [FiftyOneImageDetectionDataset](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlfiftyoneimagedetectiondataset) | β | β | β || [COCODetectionDataset](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlcocodetectiondataset) | β | β | β || [VOCDetectionDataset](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlvocdetectiondataset) | β | β | β || [KITTIDetectionDataset](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlkittidetectiondataset) | β | β | β || [YOLODataset](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlyolodataset) | β | β | β || [TFObjectDetectionDataset](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmltfobjectdetectiondataset) | β | β | β || [CVATImageDataset](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlcvatimagedataset) | β | β | β || [CVATVideoDataset](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlcvatvideodataset) | β | β | β || [FiftyOneImageLabelsDataset](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlfiftyoneimagelabelsdataset) | β | β | β || [FiftyOneVideoLabelsDataset](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlfiftyonevideolabelsdataset) | β | β | β || [BDDDataset](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlbdddataset) | β | β | β | In addition, you can define your own [custom dataset types](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlcustom-formats) to read/write datasets in your own formats.The usage of the `fiftyone convert` command is as follows: | fiftyone convert -h | usage: fiftyone convert [-h] [--input-dir INPUT_DIR] [--input-type INPUT_TYPE]
[--output-dir OUTPUT_DIR] [--output-type OUTPUT_TYPE]
Convert datasets on disk between supported formats.
Examples::
# Convert an image classification directory tree to TFRecords format
fiftyone convert \
--input-dir /path/to/image-classification-directory-tree \
--input-type fiftyone.types.ImageClassificationDirectoryTree \
--output-dir /path/for/tf-image-classification-dataset \
--output-type fiftyone.types.TFImageClassificationDataset
# Convert a COCO detection dataset to CVAT image format
fiftyone convert \
--input-dir /path/to/coco-detection-dataset \
--input-type fiftyone.types.COCODetectionDataset \
--output-dir /path/for/cvat-image-dataset \
--output-type fiftyone.types.CVATImageDataset
optional arguments:
-h, --help show this help message and exit
--input-dir INPUT_DIR
the directory containing the dataset
--input-type INPUT_TYPE
the fiftyone.types.Dataset type of the input dataset
--output-dir OUTPUT_DIR
the directory to which to write the output dataset
--output-type OUTPUT_TYPE
the fiftyone.types.Dataset type to output
| Apache-2.0 | docs/source/recipes/convert_datasets.ipynb | pixta-dev/fiftyone |
Convert CIFAR-10 dataset When you downloaded the test split of the CIFAR-10 dataset above, it was written to disk as a dataset in [fiftyone.types.FiftyOneImageClassificationDataset](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlfiftyoneimageclassificationdataset) format.You can verify this by printing information about the downloaded dataset: | fiftyone zoo datasets info cifar10 | ***** Dataset description *****
The CIFAR-10 dataset consists of 60000 32 x 32 color images in 10
classes, with 6000 images per class. There are 50000 training images and
10000 test images.
Dataset size:
132.40 MiB
Source:
https://www.cs.toronto.edu/~kriz/cifar.html
***** Supported splits *****
test, train
***** Dataset location *****
~/fiftyone/cifar10
***** Dataset info *****
{
"name": "cifar10",
"zoo_dataset": "fiftyone.zoo.datasets.torch.CIFAR10Dataset",
"dataset_type": "fiftyone.types.dataset_types.FiftyOneImageClassificationDataset",
"num_samples": 10000,
"downloaded_splits": {
"test": {
"split": "test",
"num_samples": 10000
}
},
"classes": [
"airplane",
"automobile",
"bird",
"cat",
"deer",
"dog",
"frog",
"horse",
"ship",
"truck"
]
}
| Apache-2.0 | docs/source/recipes/convert_datasets.ipynb | pixta-dev/fiftyone |
The snippet below uses `fiftyone convert` to convert the test split of the CIFAR-10 dataset to [fiftyone.types.ImageClassificationDirectoryTree](https://voxel51.com/docs/fiftyone/user_guide/export_datasets.htmlimageclassificationdirectorytree) format, which stores classification datasets on disk in a directory tree structure with images organized per-class:```βββ /β βββ .β βββ .β βββ ...βββ /β βββ .β βββ .β βββ ...βββ ...``` | INPUT_DIR=$(fiftyone zoo datasets find cifar10 --split test)
OUTPUT_DIR=/tmp/fiftyone/cifar10-dir-tree
fiftyone convert \
--input-dir ${INPUT_DIR} --input-type fiftyone.types.FiftyOneImageClassificationDataset \
--output-dir ${OUTPUT_DIR} --output-type fiftyone.types.ImageClassificationDirectoryTree | Loading dataset from '~/fiftyone/cifar10/test'
Input format 'fiftyone.types.dataset_types.FiftyOneImageClassificationDataset'
100% |βββ| 10000/10000 [4.2s elapsed, 0s remaining, 2.4K samples/s]
Import complete
Exporting dataset to '/tmp/fiftyone/cifar10-dir-tree'
Export format 'fiftyone.types.dataset_types.ImageClassificationDirectoryTree'
100% |βββ| 10000/10000 [6.2s elapsed, 0s remaining, 1.7K samples/s]
Export complete
| Apache-2.0 | docs/source/recipes/convert_datasets.ipynb | pixta-dev/fiftyone |
Let's verify that the conversion happened as expected: | ls -lah /tmp/fiftyone/cifar10-dir-tree/
ls -lah /tmp/fiftyone/cifar10-dir-tree/airplane/ | head | total 8000
drwxr-xr-x 1002 voxel51 wheel 31K Jul 14 11:08 .
drwxr-xr-x 12 voxel51 wheel 384B Jul 14 11:08 ..
-rw-r--r-- 1 voxel51 wheel 1.2K Jul 14 11:23 000004.jpg
-rw-r--r-- 1 voxel51 wheel 1.1K Jul 14 11:23 000011.jpg
-rw-r--r-- 1 voxel51 wheel 1.1K Jul 14 11:23 000022.jpg
-rw-r--r-- 1 voxel51 wheel 1.3K Jul 14 11:23 000028.jpg
-rw-r--r-- 1 voxel51 wheel 1.2K Jul 14 11:23 000045.jpg
-rw-r--r-- 1 voxel51 wheel 1.2K Jul 14 11:23 000053.jpg
-rw-r--r-- 1 voxel51 wheel 1.3K Jul 14 11:23 000075.jpg
| Apache-2.0 | docs/source/recipes/convert_datasets.ipynb | pixta-dev/fiftyone |
Now let's convert the classification directory tree to [TFRecords](https://voxel51.com/docs/fiftyone/user_guide/export_datasets.htmltfimageclassificationdataset) format! | INPUT_DIR=/tmp/fiftyone/cifar10-dir-tree
OUTPUT_DIR=/tmp/fiftyone/cifar10-tfrecords
fiftyone convert \
--input-dir ${INPUT_DIR} --input-type fiftyone.types.ImageClassificationDirectoryTree \
--output-dir ${OUTPUT_DIR} --output-type fiftyone.types.TFImageClassificationDataset | Loading dataset from '/tmp/fiftyone/cifar10-dir-tree'
Input format 'fiftyone.types.dataset_types.ImageClassificationDirectoryTree'
100% |βββ| 10000/10000 [4.0s elapsed, 0s remaining, 2.5K samples/s]
Import complete
Exporting dataset to '/tmp/fiftyone/cifar10-tfrecords'
Export format 'fiftyone.types.dataset_types.TFImageClassificationDataset'
0% ||--| 1/10000 [23.2ms elapsed, 3.9m remaining, 43.2 samples/s] 2020-07-14 11:24:15.187387: I tensorflow/core/platform/cpu_feature_guard.cc:143] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA
2020-07-14 11:24:15.201384: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x7f83df428f60 initialized for platform Host (this does not guarantee that XLA will be used). Devices:
2020-07-14 11:24:15.201405: I tensorflow/compiler/xla/service/service.cc:176] StreamExecutor device (0): Host, Default Version
100% |βββ| 10000/10000 [8.2s elapsed, 0s remaining, 1.3K samples/s]
Export complete
| Apache-2.0 | docs/source/recipes/convert_datasets.ipynb | pixta-dev/fiftyone |
Let's verify that the conversion happened as expected: | ls -lah /tmp/fiftyone/cifar10-tfrecords | total 29696
drwxr-xr-x 3 voxel51 wheel 96B Jul 14 11:24 .
drwxr-xr-x 4 voxel51 wheel 128B Jul 14 11:24 ..
-rw-r--r-- 1 voxel51 wheel 14M Jul 14 11:24 tf.records
| Apache-2.0 | docs/source/recipes/convert_datasets.ipynb | pixta-dev/fiftyone |
Convert KITTI dataset When you downloaded the validation split of the KITTI dataset above, it was written to disk as a dataset in [fiftyone.types.FiftyOneImageDetectionDataset](https://voxel51.com/docs/fiftyone/user_guide/dataset_creation/datasets.htmlfiftyoneimagedetectiondataset) format.You can verify this by printing information about the downloaded dataset: | fiftyone zoo datasets info kitti | ***** Dataset description *****
KITTI contains a suite of vision tasks built using an autonomous
driving platform.
The full benchmark contains many tasks such as stereo, optical flow, visual
odometry, etc. This dataset contains the object detection dataset,
including the monocular images and bounding boxes. The dataset contains
7481 training images annotated with 3D bounding boxes. A full description
of the annotations can be found in the README of the object development kit
on the KITTI homepage.
Dataset size:
5.27 GiB
Source:
http://www.cvlibs.net/datasets/kitti
***** Supported splits *****
test, train, validation
***** Dataset location *****
~/fiftyone/kitti
***** Dataset info *****
{
"name": "kitti",
"zoo_dataset": "fiftyone.zoo.datasets.tf.KITTIDataset",
"dataset_type": "fiftyone.types.dataset_types.FiftyOneImageDetectionDataset",
"num_samples": 423,
"downloaded_splits": {
"validation": {
"split": "validation",
"num_samples": 423
}
},
"classes": [
"Car",
"Van",
"Truck",
"Pedestrian",
"Person_sitting",
"Cyclist",
"Tram",
"Misc"
]
}
| Apache-2.0 | docs/source/recipes/convert_datasets.ipynb | pixta-dev/fiftyone |
The snippet below uses `fiftyone convert` to convert the test split of the CIFAR-10 dataset to [fiftyone.types.COCODetectionDataset](https://voxel51.com/docs/fiftyone/user_guide/export_datasets.htmlcocodetectiondataset) format, which writes the dataset to disk with annotations in [COCO format](https://cocodataset.org/format-data). | INPUT_DIR=$(fiftyone zoo datasets find kitti --split validation)
OUTPUT_DIR=/tmp/fiftyone/kitti-coco
fiftyone convert \
--input-dir ${INPUT_DIR} --input-type fiftyone.types.FiftyOneImageDetectionDataset \
--output-dir ${OUTPUT_DIR} --output-type fiftyone.types.COCODetectionDataset | Loading dataset from '~/fiftyone/kitti/validation'
Input format 'fiftyone.types.dataset_types.FiftyOneImageDetectionDataset'
100% |βββββββ| 423/423 [1.2s elapsed, 0s remaining, 351.0 samples/s]
Import complete
Exporting dataset to '/tmp/fiftyone/kitti-coco'
Export format 'fiftyone.types.dataset_types.COCODetectionDataset'
100% |βββββββ| 423/423 [4.4s elapsed, 0s remaining, 96.1 samples/s]
Export complete
| Apache-2.0 | docs/source/recipes/convert_datasets.ipynb | pixta-dev/fiftyone |
Let's verify that the conversion happened as expected: | ls -lah /tmp/fiftyone/kitti-coco/
ls -lah /tmp/fiftyone/kitti-coco/data | head
cat /tmp/fiftyone/kitti-coco/labels.json | python -m json.tool 2> /dev/null | head -20
echo "..."
cat /tmp/fiftyone/kitti-coco/labels.json | python -m json.tool 2> /dev/null | tail -20 | {
"info": {
"year": "",
"version": "",
"description": "Exported from FiftyOne",
"contributor": "",
"url": "https://voxel51.com/fiftyone",
"date_created": "2020-07-14T11:24:40"
},
"licenses": [],
"categories": [
{
"id": 0,
"name": "Car",
"supercategory": "none"
},
{
"id": 1,
"name": "Cyclist",
"supercategory": "none"
...
"area": 4545.8,
"segmentation": null,
"iscrowd": 0
},
{
"id": 3196,
"image_id": 422,
"category_id": 3,
"bbox": [
367.2,
107.3,
36.2,
105.2
],
"area": 3808.2,
"segmentation": null,
"iscrowd": 0
}
]
}
| Apache-2.0 | docs/source/recipes/convert_datasets.ipynb | pixta-dev/fiftyone |
Now let's convert from COCO format to [CVAT Image format](https://voxel51.com/docs/fiftyone/user_guide/export_datasets.htmlcvatimageformat) format! | INPUT_DIR=/tmp/fiftyone/kitti-coco
OUTPUT_DIR=/tmp/fiftyone/kitti-cvat
fiftyone convert \
--input-dir ${INPUT_DIR} --input-type fiftyone.types.COCODetectionDataset \
--output-dir ${OUTPUT_DIR} --output-type fiftyone.types.CVATImageDataset | Loading dataset from '/tmp/fiftyone/kitti-coco'
Input format 'fiftyone.types.dataset_types.COCODetectionDataset'
100% |βββββββ| 423/423 [2.0s elapsed, 0s remaining, 206.4 samples/s]
Import complete
Exporting dataset to '/tmp/fiftyone/kitti-cvat'
Export format 'fiftyone.types.dataset_types.CVATImageDataset'
100% |βββββββ| 423/423 [1.3s elapsed, 0s remaining, 323.7 samples/s]
Export complete
| Apache-2.0 | docs/source/recipes/convert_datasets.ipynb | pixta-dev/fiftyone |
Let's verify that the conversion happened as expected: | ls -lah /tmp/fiftyone/kitti-cvat
cat /tmp/fiftyone/kitti-cvat/labels.xml | head -20
echo "..."
cat /tmp/fiftyone/kitti-cvat/labels.xml | tail -20 | <?xml version="1.0" encoding="utf-8"?>
<annotations>
<version>1.1</version>
<meta>
<task>
<size>423</size>
<mode>annotation</mode>
<labels>
<label>
<name>Car</name>
<attributes>
</attributes>
</label>
<label>
<name>Cyclist</name>
<attributes>
</attributes>
</label>
<label>
<name>Misc</name>
...
<box label="Pedestrian" xtl="360" ytl="116" xbr="402" ybr="212">
</box>
<box label="Pedestrian" xtl="396" ytl="120" xbr="430" ybr="212">
</box>
<box label="Pedestrian" xtl="413" ytl="112" xbr="483" ybr="212">
</box>
<box label="Pedestrian" xtl="585" ytl="80" xbr="646" ybr="215">
</box>
<box label="Pedestrian" xtl="635" ytl="94" xbr="688" ybr="212">
</box>
<box label="Pedestrian" xtl="422" ytl="85" xbr="469" ybr="210">
</box>
<box label="Pedestrian" xtl="457" ytl="93" xbr="520" ybr="213">
</box>
<box label="Pedestrian" xtl="505" ytl="101" xbr="548" ybr="206">
</box>
<box label="Pedestrian" xtl="367" ytl="107" xbr="403" ybr="212">
</box>
</image>
</annotations> | Apache-2.0 | docs/source/recipes/convert_datasets.ipynb | pixta-dev/fiftyone |
CleanupYou can cleanup the files generated by this recipe by running the command below: | rm -rf /tmp/fiftyone | _____no_output_____ | Apache-2.0 | docs/source/recipes/convert_datasets.ipynb | pixta-dev/fiftyone |
Session 1 Homework Solution=========================This is the homework for the first session of the MolSSI Python Scripting Level 2 Workshop.This homework is intended to give you practice with the material covered in the first session. Goals: - Utilize pandas to read in and work with the data in a csv file. - Utilize matplotlib to create plots and subplots of data stored in a pandas dataframe. - Utilize pandas to extract specific information from a dataframe. - Utilize matplotlib to construct a plot with multiple groups of data. Exercise 1Using the `PubChemElements_all.csv` file used during Session 1, create a plot of the Ionization Energy trend of the periodic table. The trend can be visualized by plotting the Ionization Energy of each element against it's Atomic Number. HINT: Use pandas to read in the csv file and plot the data using matplotlib. Exercise 2Create a pair of subplots comparing the Ionization Energy and Electronegativity trends. The Electronegativity trend can be plotted in the same way as the Ionization Energy, by plotting the Electronegativity of each element against it's Atomic Number. Exercise 3 Create a function that will assign a color coding to a particular Standard State of an element, i.e. red for gases, blue for solids, etc. Use the apply function from pandas to apply the function across the periodic table, creating a column of assigned colors. Create a pair of subplots utilizing the assigned color as the color of the marker: - Atomic Mass against the Melting point of each element - Ionization Energy against the Melting Point of each element Exercise 1 Solution | # Import necessary packages for the homework:
import os
import pandas as pd
import matplotlib.pyplot as plt
# %matplotlib notebook
# Create a filepath to the periodic table csv file.
file_path = os.path.join("data", "PubChemElements_all.csv")
# Use pandas to read the csv file into a table.
df = pd.read_csv(file_path)
# Print the first 5 rows of the table to get a quick glance at its contents.
df.head()
# Create a simple scatter plot of the Ionization Energy vs the Atomic number of each element.
ion_fig, ion_ax = plt.subplots()
ion_ax.scatter('AtomicNumber', 'IonizationEnergy', data=df)
ion_ax.set_xlabel('Atomic Number')
ion_ax.set_ylabel('Ionization Energy') | _____no_output_____ | BSD-3-Clause | book/homework_1_solutions.ipynb | janash/python-analysis |
Exercise 2 Solution | # Create a set of subplots of the two trends: Ionization Energy and Electronegativity.
comparison_fig, comparison_ax = plt.subplots(1, 2)
# Add the first subplot.
comparison_ax[0].scatter('AtomicNumber', 'IonizationEnergy', data=df)
comparison_ax[0].set_xlabel('Atomic Number')
comparison_ax[0].set_ylabel('Ionization')
comparison_fig.tight_layout()
# Add the second subplot.
comparison_ax[1].scatter('AtomicNumber', 'Electronegativity', data=df)
comparison_ax[1].set_xlabel('Atomic Number')
comparison_ax[1].set_ylabel('Electronegativity')
comparison_fig.tight_layout()
comparison_fig | _____no_output_____ | BSD-3-Clause | book/homework_1_solutions.ipynb | janash/python-analysis |
Exercise 3 Solution | # Determine possible states stored in the Dataframe
states = pd.unique(df['StandardState'])
states
# Create a function that returns a different color for each type of standard state.
def assign_state_color(standard_state):
state_markers = {'Gas': 'r',
'Solid': 'b',
'Liquid': 'g',
'Expected to be a Solid': 'y',
'Expected to be a Gas': 'k'}
return state_markers[standard_state]
# Apply the function to the dataframe, creating a new column.
df['StateMarker'] = df['StandardState'].apply(assign_state_color)
# Create a plot that uses the colors assigned to the table of the Atomic Mass vs the Melting Point.
state_fig, state_ax = plt.subplots(1, 2, figsize=(8,4))
state_ax[0].set_xlabel('Atomic Mass')
state_ax[0].set_ylabel('Melting Point')
# Add each state as a separate scatter to the same subplot.
for state in states:
dataframe = df[df['StandardState'] == state]
line = state_ax[0].scatter('AtomicMass', 'MeltingPoint', data=dataframe, color=dataframe.iloc[0]['StateMarker'])
line.set_label(state)
state_ax[1].set_xlabel('Ionization Energy')
state_ax[1].set_ylabel('Melting Point')
for state in states:
dataframe = df[df['StandardState'] == state]
line = state_ax[1].scatter('IonizationEnergy', 'MeltingPoint', data=dataframe, color=dataframe.iloc[0]['StateMarker'])
line.set_label(state)
# Create a legend
state_ax[1].legend(states)
state_fig.tight_layout() | _____no_output_____ | BSD-3-Clause | book/homework_1_solutions.ipynb | janash/python-analysis |
Generator function to armstrong numbers | def genFunc():
start = 1
end = 1000
for i in range(start, end + 1):
if i >= 10:
order = len(str(i))
sum = 0
temp = i
while temp > 0:
dig = temp % 10
sum += dig ** order
temp //= 10
if i == sum:
yield i
for x in genFunc():
print(x) | 153
370
371
407
| Apache-2.0 | Assignment Day 9 q2.ipynb | gopi2650/letsupgrade-python |
define categorical / numeric columns | d.hist(figsize=[20, 20], sharey=False, bins=50)
d.hist(figsize=[20, 20], sharey=True)
print()
col_identity = {'ignore': ['accident_id','provider_and_id','provider_code'],
'numeric' : ['license_acquiring_date', 'accident_year','accident_month'],
'category' : ['age_group', 'sex', 'vehicle_type', 'safety_measures', 'population_type', 'home_region', 'home_district', 'home_natural_area', 'home_municipal_status', 'home_residence_type', 'medical_type', 'safety_measures_use', 'car_id', 'involve_id']}
col_target = 'late_deceased'
if not 'data' in globals():
data_fname = 'data/involved_hebrew_dummies.parquet'
if os.path.isfile(data_fname):
data = pd.read_parquet(data_fname)
else:
rel_cols = col_identity['category'] + col_identity['numeric']
data = d[col_identity['numeric']].fillna(-1)
data['license_acquiring_date'] = data['license_acquiring_date'].replace(0, -1)
dummies = pd.get_dummies(d[col_identity['category']], columns=col_identity['category'], prefix_sep='==')
data = pd.concat([dummies, data], axis=1)
data.to_parquet(data_fname)
data.head()
y = d[col_target].fillna(0)
y[y == 2] = 1
y = y.astype(np.int16).values
x = data.fillna(-1).values
y.sum(), y.shape[0], x.shape | _____no_output_____ | MIT | datascience/2018_10_27_anyway_data_trial_1.ipynb | neuhofmo/anyway_projects |
balance the data and train forest of decision trees | from sklearn.model_selection import train_test_split
from imblearn.ensemble import BalancedRandomForestClassifier # !!!!!! balanced!
from sklearn.metrics import f1_score, precision_score, recall_score, balanced_accuracy_score
del df
X_train, X_test, y_train, y_test = train_test_split(x, y, random_state=42, test_size=0.25)
print(X_train.shape, X_test.shape, y_train.shape, y_test.shape)
print("%.4f, %.4f"%(y_train.sum() / len(y_train), y_test.sum() / len(y_test)))
brf = BalancedRandomForestClassifier(n_estimators=30, random_state=0)
brf.fit(X_train, y_train)
y_pred = brf.predict(X_test)
print('f1 score = %.3f'%(f1_score(y_test, y_pred, average='weighted')))
print('precision = %.3f'%(precision_score(y_test, y_pred, average='weighted')))
print('recall = %.3f'%(recall_score(y_test, y_pred, average='weighted')))
print('accuracy(balanced) = %.3f'%(balanced_accuracy_score(y_test, y_pred)))
importance = pd.DataFrame(list(zip(data.columns, brf.feature_importances_)), columns=['feature', 'importance']).sort_values('importance', ascending=False)
importance[0:20].style.bar()
if not os.path.isdir('models'): os.mkdir('models')
from sklearn.externals import joblib
joblib.dump(brf, 'models/2018_10_28_death_risk_balanced_RF_classifier_01.joblib')
import lime
#eplain?
#see http://explained.ai/rf-importance/index.html
from sklearn.metrics import f1_score, precision_score, recall_score, balanced_accuracy_score
def metric(model, x, y):
return balanced_accuracy_score(y, model.predict(x))
def permutation_importances(model, X_train, y_train):
baseline = metric(model, X_train, y_train)
imp = []
for col in X_train.columns:
save = X_train[col].copy()
X_train[col] = np.random.permutation(X_train[col])
m = metric(model, X_train, y_train)
X_train[col] = save
imp.append(baseline - m)
return np.array(imp)
#permutation_importances(brf, data, y_train)
d.groupby('accident_year').count()['accident_id'].plot(kind='bar') | _____no_output_____ | MIT | datascience/2018_10_27_anyway_data_trial_1.ipynb | neuhofmo/anyway_projects |
Data Support | fan_smoothing_window = 60 # time width of smoothing wndow
def load_df(df_name):
df = pd.read_csv(
df_name,
usecols=[0, 1, 2, 3, 4, 5, 6]
)
df.rename(index=str, columns={ # remove units for easier indexing
'Time (s)': 'Time',
'Temperature': 'Thermistor',
'Error (deg C)': 'Error',
'Setpoint Reached': 'Reached'
}, inplace=True)
df.dropna(how='all', inplace=True)
df.index = pd.to_timedelta(df['Time'], unit='s') # set index to units of seconds
df.Time = df.Time / 60 # set Time to units of minutes
return df
def smooth_fan(df):
df['FanSmooth'] = df.Fan.rolling(fan_smoothing_window, win_type='hamming').mean() | _____no_output_____ | BSD-3-Clause | results/Temperature Control Plots.ipynb | ethanjli/punchcard-microfluidics |
Plotting Support | figure_width = 17.5
figure_temps_height = 4
figure_complete_height = 7.5
figure_complete_height_ratio = (3, 2)
box_width_shrink_factor = 0.875 # to fit the figure legend on the right
ylabel_position = -0.08
min_temp = 20
max_temp = 100
legend_location = 'center right'
reached_color = 'gainsboro' # light gray
setpoint_color = 'tab:green'
thermistor_color = 'tab:orange'
fan_color = 'tab:blue'
heater_color = 'tab:red'
def fig_temps(title):
(fig, ax_temp) = plt.subplots(
figsize=(figure_width, figure_temps_height)
)
ax_temp.set_title(title)
return (fig, ax_temp)
def fig_complete(title):
(fig, (ax_temp, ax_duties)) = plt.subplots(
nrows=2, sharex=True,
gridspec_kw={
'height_ratios': figure_complete_height_ratio
},
figsize=(figure_width, figure_complete_height)
)
ax_temp.set_title(title)
return (fig, (ax_temp, ax_duties))
def plot_setpoint_reached(df, ax, label=True):
legend_label = 'Reached\nSetpoint'
if not label:
legend_label = '_' + legend_label # hide the label from the legend
(groups, _) = ndi.label(df.Reached.values.tolist())
df = pd.DataFrame({
'Time': df.Time,
'ReachedGroup': groups
})
result = (
df
.loc[df.ReachedGroup != 0]
.groupby('ReachedGroup')['Time']
.agg(['first', 'last'])
)
for (i, (group_start, group_end)) in enumerate(result.values.tolist()):
ax.axvspan(group_start, group_end, facecolor=reached_color, label=legend_label)
if i == 0:
legend_label = '_' + legend_label # hide subsequent labels from the legend
def plot_temps(df, ax, label_x=True):
ax.plot(df.Time, df.Setpoint, color=setpoint_color, label='Setpoint')
ax.plot(df.Time, df.Thermistor, color=thermistor_color, label='Thermistor')
ax.set_xlim([df.Time[0], df.Time[-1]])
ax.set_ylim([min_temp, max_temp])
if label_x:
ax.set_xlabel('Time (min)')
ax.set_ylabel('Temperature\n(Β°C)')
def plot_efforts(df, ax):
ax.plot(df.Time, df.FanSmooth, color=fan_color, label='Fan')
ax.plot(df.Time, df.Heater, color=heater_color, label='Heater')
ax.set_xlabel('Time (min)')
ax.set_ylabel('Duty\nCycle')
def shrink_ax_width(ax, shrink_factor):
box = ax.get_position()
ax.set_position([box.x0, box.y0, box.width * shrink_factor, box.height])
def fig_plot_complete(df, title):
(fig, (ax_temp, ax_duties)) = fig_complete(title)
plot_setpoint_reached(df, ax_temp)
plot_temps(df, ax_temp, label_x=False)
ax_temp.yaxis.set_label_coords(ylabel_position, 0.5)
shrink_ax_width(ax_temp, box_width_shrink_factor)
plot_setpoint_reached(df, ax_duties, label=False)
plot_efforts(df, ax_duties)
ax_duties.yaxis.set_label_coords(ylabel_position, 0.5)
shrink_ax_width(ax_duties, box_width_shrink_factor)
fig.legend(loc=legend_location)
def fig_plot_temps(df, title):
(fig, ax_temp) = fig_temps(title)
plot_setpoint_reached(df, ax_temp)
plot_temps(df, ax_temp)
ax_temp.yaxis.set_label_coords(ylabel_position, 0.5)
shrink_ax_width(ax_temp, box_width_shrink_factor)
fig.legend(loc=legend_location) | _____no_output_____ | BSD-3-Clause | results/Temperature Control Plots.ipynb | ethanjli/punchcard-microfluidics |
Stepwise Sequence | df_stepwise = load_df('20190117 Thermal Subsystem Testing Data - Fifth Test.csv')
smooth_fan(df_stepwise)
df_stepwise
fig_plot_temps(df_stepwise, 'Stepwise Adjustment Control Sequence')
plt.savefig('stepwise_control.pdf', format='pdf')
plt.savefig('stepwise_control.png', format='png')
fig_plot_complete(df_stepwise, 'Stepwise Control Sequence') | _____no_output_____ | BSD-3-Clause | results/Temperature Control Plots.ipynb | ethanjli/punchcard-microfluidics |
Lysis Sequence | df_lysis = load_df('20190117 Thermal Subsystem Testing Data - Fourth Test.csv')
smooth_fan(df_lysis)
df_lysis
fig_plot_temps(df_lysis, 'Thermal Lysis Control Sequence')
fig_plot_complete(df_lysis, 'Thermal Lysis Control Sequence')
plt.savefig('thermal_lysis.pdf', format='pdf')
plt.savefig('thermal_lysis.png', format='png') | 'HelveticaNeueLTStd_Md.otf' can not be subsetted into a Type 3 font. The entire font will be embedded in the output.
'HelveticaNeueLTStd_Roman.otf' can not be subsetted into a Type 3 font. The entire font will be embedded in the output.
'HelveticaNeueLTStd_Md.otf' can not be subsetted into a Type 3 font. The entire font will be embedded in the output.
'HelveticaNeueLTStd_Roman.otf' can not be subsetted into a Type 3 font. The entire font will be embedded in the output.
| BSD-3-Clause | results/Temperature Control Plots.ipynb | ethanjli/punchcard-microfluidics |
**Table of Contents:*** Introduction* The RMS Titanic* Import Libraries* Getting the Data* Data Exploration/Analysis* Data Preprocessing - Missing Data - Converting Features - Creating Categories - Creating new Features* Building Machine Learning Models - Training 8 different models - Which is the best model ? - K-Fold Cross Validation* Random Forest - What is Random Forest ? - Feature importance - Hyperparameter Tuning * Further Evaluation - Confusion Matrix - Precision and Recall - F-Score - Precision Recall Curve - ROC AUC Curve - ROC AUC Score* Submission* Summary **Introduction**In this kernel I will go through the whole process of creating a machine learning model on the famous Titanic dataset, which is used by many people all over the world. It provides information on the fate of passengers on the Titanic, summarized according to economic status (class), sex, age and survival. In this challenge, we are asked to predict whether a passenger on the titanic would have been survived or not. **The RMS Titanic**RMS Titanic was a British passenger liner that sank in the North Atlantic Ocean in the early morning hours of 15 April 1912, after it collided with an iceberg during its maiden voyage from Southampton to New York City. There were an estimated 2,224 passengers and crew aboard the ship, and more than 1,500 died, making it one of the deadliest commercial peacetime maritime disasters in modern history. The RMS Titanic was the largest ship afloat at the time it entered service and was the second of three Olympic-class ocean liners operated by the White Star Line. The Titanic was built by the Harland and Wolff shipyard in Belfast. Thomas Andrews, her architect, died in the disaster.  **Import Libraries** | # linear algebra
import numpy as np
# data processing
import pandas as pd
# data visualization
import seaborn as sns
%matplotlib inline
from matplotlib import pyplot as plt
from matplotlib import style
# Algorithms
from sklearn import linear_model
from sklearn.linear_model import LogisticRegression
from sklearn.ensemble import RandomForestClassifier
from sklearn.linear_model import Perceptron
from sklearn.linear_model import SGDClassifier
from sklearn.tree import DecisionTreeClassifier
from sklearn.neighbors import KNeighborsClassifier
from sklearn.svm import SVC, LinearSVC
from sklearn.naive_bayes import GaussianNB | _____no_output_____ | MIT | titanic/end-to-end-project-with-python.ipynb | MLVPRASAD/KaggleProjects |
**Getting the Data** | test_df = pd.read_csv("../input/test.csv")
train_df = pd.read_csv("../input/train.csv") | _____no_output_____ | MIT | titanic/end-to-end-project-with-python.ipynb | MLVPRASAD/KaggleProjects |
**Data Exploration/Analysis** | train_df.info() | <class 'pandas.core.frame.DataFrame'>
RangeIndex: 891 entries, 0 to 890
Data columns (total 12 columns):
PassengerId 891 non-null int64
Survived 891 non-null int64
Pclass 891 non-null int64
Name 891 non-null object
Sex 891 non-null object
Age 714 non-null float64
SibSp 891 non-null int64
Parch 891 non-null int64
Ticket 891 non-null object
Fare 891 non-null float64
Cabin 204 non-null object
Embarked 889 non-null object
dtypes: float64(2), int64(5), object(5)
memory usage: 83.6+ KB
| MIT | titanic/end-to-end-project-with-python.ipynb | MLVPRASAD/KaggleProjects |
**The training-set has 891 examples and 11 features + the target variable (survived)**. 2 of the features are floats, 5 are integers and 5 are objects. Below I have listed the features with a short description: survival: Survival PassengerId: Unique Id of a passenger. pclass: Ticket class sex: Sex Age: Age in years sibsp: of siblings / spouses aboard the Titanic parch: of parents / children aboard the Titanic ticket: Ticket number fare: Passenger fare cabin: Cabin number embarked: Port of Embarkation | train_df.describe() | _____no_output_____ | MIT | titanic/end-to-end-project-with-python.ipynb | MLVPRASAD/KaggleProjects |
Above we can see that **38% out of the training-set survived the Titanic**. We can also see that the passenger ages range from 0.4 to 80. On top of that we can already detect some features, that contain missing values, like the 'Age' feature. | train_df.head(15) | _____no_output_____ | MIT | titanic/end-to-end-project-with-python.ipynb | MLVPRASAD/KaggleProjects |
From the table above, we can note a few things. First of all, that we **need to convert a lot of features into numeric** ones later on, so that the machine learning algorithms can process them. Furthermore, we can see that the **features have widely different ranges**, that we will need to convert into roughly the same scale. We can also spot some more features, that contain missing values (NaN = not a number), that wee need to deal with.**Let's take a more detailed look at what data is actually missing:** | total = train_df.isnull().sum().sort_values(ascending=False)
percent_1 = train_df.isnull().sum()/train_df.isnull().count()*100
percent_2 = (round(percent_1, 1)).sort_values(ascending=False)
missing_data = pd.concat([total, percent_2], axis=1, keys=['Total', '%'])
missing_data.head(5) | _____no_output_____ | MIT | titanic/end-to-end-project-with-python.ipynb | MLVPRASAD/KaggleProjects |
The Embarked feature has only 2 missing values, which can easily be filled. It will be much more tricky, to deal with the 'Age' feature, which has 177 missing values. The 'Cabin' feature needs further investigation, but it looks like that we might want to drop it from the dataset, since 77 % of it are missing. | train_df.columns.values | _____no_output_____ | MIT | titanic/end-to-end-project-with-python.ipynb | MLVPRASAD/KaggleProjects |
Above you can see the 11 features + the target variable (survived). **What features could contribute to a high survival rate ?** To me it would make sense if everything except 'PassengerId', 'Ticket' and 'Name' would be correlated with a high survival rate. **1. Age and Sex:** | survived = 'survived'
not_survived = 'not survived'
fig, axes = plt.subplots(nrows=1, ncols=2,figsize=(10, 4))
women = train_df[train_df['Sex']=='female']
men = train_df[train_df['Sex']=='male']
ax = sns.distplot(women[women['Survived']==1].Age.dropna(), bins=18, label = survived, ax = axes[0], kde =False)
ax = sns.distplot(women[women['Survived']==0].Age.dropna(), bins=40, label = not_survived, ax = axes[0], kde =False)
ax.legend()
ax.set_title('Female')
ax = sns.distplot(men[men['Survived']==1].Age.dropna(), bins=18, label = survived, ax = axes[1], kde = False)
ax = sns.distplot(men[men['Survived']==0].Age.dropna(), bins=40, label = not_survived, ax = axes[1], kde = False)
ax.legend()
_ = ax.set_title('Male') | _____no_output_____ | MIT | titanic/end-to-end-project-with-python.ipynb | MLVPRASAD/KaggleProjects |
You can see that men have a high probability of survival when they are between 18 and 30 years old, which is also a little bit true for women but not fully. For women the survival chances are higher between 14 and 40.For men the probability of survival is very low between the age of 5 and 18, but that isn't true for women. Another thing to note is that infants also have a little bit higher probability of survival.Since there seem to be **certain ages, which have increased odds of survival** and because I want every feature to be roughly on the same scale, I will create age groups later on. **3. Embarked, Pclass and Sex:** | FacetGrid = sns.FacetGrid(train_df, row='Embarked', size=4.5, aspect=1.6)
FacetGrid.map(sns.pointplot, 'Pclass', 'Survived', 'Sex', palette=None, order=None, hue_order=None )
FacetGrid.add_legend() | /opt/conda/lib/python3.6/site-packages/seaborn/axisgrid.py:230: UserWarning: The `size` paramter has been renamed to `height`; please update your code.
warnings.warn(msg, UserWarning)
/opt/conda/lib/python3.6/site-packages/scipy/stats/stats.py:1713: FutureWarning: Using a non-tuple sequence for multidimensional indexing is deprecated; use `arr[tuple(seq)]` instead of `arr[seq]`. In the future this will be interpreted as an array index, `arr[np.array(seq)]`, which will result either in an error or a different result.
return np.add.reduce(sorted[indexer] * weights, axis=axis) / sumval
| MIT | titanic/end-to-end-project-with-python.ipynb | MLVPRASAD/KaggleProjects |
Embarked seems to be correlated with survival, depending on the gender. Women on port Q and on port S have a higher chance of survival. The inverse is true, if they are at port C. Men have a high survival probability if they are on port C, but a low probability if they are on port Q or S. Pclass also seems to be correlated with survival. We will generate another plot of it below. **4. Pclass:** | sns.barplot(x='Pclass', y='Survived', data=train_df) | /opt/conda/lib/python3.6/site-packages/scipy/stats/stats.py:1713: FutureWarning: Using a non-tuple sequence for multidimensional indexing is deprecated; use `arr[tuple(seq)]` instead of `arr[seq]`. In the future this will be interpreted as an array index, `arr[np.array(seq)]`, which will result either in an error or a different result.
return np.add.reduce(sorted[indexer] * weights, axis=axis) / sumval
| MIT | titanic/end-to-end-project-with-python.ipynb | MLVPRASAD/KaggleProjects |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.