ID
int64 1
1.07k
| Comment
stringlengths 8
1.13k
| Code
stringlengths 10
4.28k
| Label
stringclasses 4
values | Source
stringlengths 21
21
| File
stringlengths 4
82
|
|---|---|---|---|---|---|
501 | function to extract the sequence "Obj_xxx" untile a period "." appears | slice_object_string <- function(string) { gsub(".*Obj_(.*)\\..*", "\\1", c(string)) } familiar_objects <- unlist(lapply(string_list, slice_object_string)) | Data Variable | https://osf.io/yfegm/ | getObjects.r |
502 | ChiSquare tests Relations between diet type and major themes | chisq.test(data$Diet, data$Health) #p = 0.94, chi-squared = 0.01 chisq.test(data$Diet, data$Food) #p = 0.44, chi-squared = 0.59 chisq.test(data$Diet, data$Social) #p = 0.80, chi-squared = 0.06 chisq.test(data$Diet, data$Logistic) #p = 0.17, chi-squared = 1.90 chisq.test(data$Diet, data$Finance) #p = 0.06, chi-squared = 3.61 chisq.test(data$Diet, data$Motivat.) #p = 0.76, chi-squared = 0.10 chisq.test(data$Diet, data$Diet.Cons.) #p = 0.05, chi-squared = 3.97 chisq.test(data$Diet, data$Other) #p = 1, chi-squared = 8.90e-29 chisq.test(data$Diet, data$Positive) #p = 0.16, chi-squared = 1.94 | Statistical Test | https://osf.io/q2zrp/ | Chi-squaretests.R |
503 | Hypotheses Fit multivariate regression model | moltenformula <- as.formula("value ~ v - 1 + v:scipopgoertz + v:age + v:sex + v:lingreg_D + v:lingreg_I + v:urbanity_log + v:edu_uni + v:edu_com + v:sciprox_score + v:sciliteracy + v:rel + v:pol + v:inter_science + v:trust_science + v:trust_scientists + v:scimedia_att + v:scimedia_sat") m_inx <- svyglm(moltenformula, design = svydsgn_melt_inx) | Statistical Modeling | https://osf.io/yhmbd/ | 02_main-analysis.R |
504 | estimate bifactor model (maximum likelihood) | fit <- cfa(model, dat,estimator = "MLM", std.lv = TRUE) | Statistical Modeling | https://osf.io/qk3bf/ | iip_estimate_bifactor.R |
505 | draw random factor scores from a multivariate normal distribution | lat_scores <- mvrnorm(n = n, mu = mu, Sigma = sigma) | Statistical Modeling | https://osf.io/qk3bf/ | iip_estimate_bifactor.R |
506 | function to create data summaries in figures | data_summary <- function(x) { m <- mean(x) ymin <- m-ci(x) ymax <- m+ci(x) return(c(y=m,ymin=ymin,ymax=ymax)) } | Visualization | https://osf.io/mj5nh/ | educationestimatecentralitydefaults.R |
507 | to get numbers for confidence intervals click on item in global environment and expand plot 95% credible intervals | x<-plot(me, plot = FALSE)[[1]] + scale_color_grey() + scale_fill_grey() p<-x+ylim(0,1)+ theme(panel.grid.major = element_line(colour="gray"), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_blank(),panel.grid.major.x = element_blank())+xlab("")+ylab("") p | Visualization | https://osf.io/a8htx/ | SSK_Cleaned_copy.R |
508 | 1 Never, 2 I tried it but don't do it systematically, 3 I do it when it feels convenient, 4 I do it for most research projects/studies, 5 I do it for every research project/study. Make dataframe 'Wish' with intentions to engage in open science practices | wish = data[, 17:25] wish_complete = wish[,-9] # remove missing data wish = apply(wish, 2, as.numeric) wish = as.data.frame(unlist(wish)) colnames(wish) = c("preregister", "sharedatas", "sharecode", "openaccess", "preprint", "openpeer", "opensw", "replicate", "other") wish = t(wish) rowMeans(wish, na.rm = T) # mean apply(wish, 1, sd, na.rm = T) # standard deviation | Data Variable | https://osf.io/zv3uc/ | analysis.R |
509 | generates 95% confidence intervals for each beta coefficient | m.stakes_decision.CI <- round(confint(m.stakes_decision, parm = "beta_"), 3) | Statistical Test | https://osf.io/uygpq/ | Cross-paradigm.R |
510 | 2.3 Loss tangents We draw a diagram of loss tangents with logarithmic scale on the xaxis | plot(graphData$Freq[graphData$site == "wreck"], graphData$tan[graphData$site == "wreck"], log = "x", axes = FALSE, xlab = "", ylab = "", type = "b", pch = PCH, col = col_wreck, xlim = c(0.01, 100), ylim = c(0, 15)) par(new = TRUE, ps = PS, mar = MAR) plot(graphData$Freq[graphData$site == "beach"], graphData$tan[graphData$site == "beach"], type = "b", log = "x", xlab = expression("Frequency [rad s"^-1 *"]"), ylab = expression(paste("tan ", delta)), pch = PCH, col = col_beach, axes = FALSE, xlim = c(0.01, 100), ylim = c(0, 15)) axis(1, at = c(0, 0.1, 1, 10, 100, 1000), labels = c(0, 0.1, 1, 10, 100, 100)) axis(2, at = c(-1 , 0, 3, 6, 9, 12, 15)) text(x = 0.01, y = 15, label = expression(bold("c"))) abline(1, 0) | Visualization | https://osf.io/9jxzs/ | 07_analysis_rheology.R |
511 | estimate mean average difference between observed and corrected values check for sign reversals | dat_analysis[sign(MA_ES_self_corr) == sign(MA_ES_self_obs), .N] # 46 (no sign rev) | Statistical Modeling | https://osf.io/dqc3y/ | REL_corr_obs_descr.R |
512 | Step 3: Fit varyingintercept ESEM model | esem_anti_fit_alt1F <- cfa(esem_anti_model_alt1F, esem_anti_data_alt, group = "country_iso", estimator = "MLM", group.equal = "loadings") | Statistical Modeling | https://osf.io/w4gey/ | 02_measures.R |
513 | performing the chisquared comparisons across all the possible pairs | pairwise.table(chi_squared_paired_comparison_1, rownames(hypotheses_per_research_area), p.adjust.method="none") pairwise.table(chi_squared_paired_comparison_2, rownames(hypotheses_per_research_area), p.adjust.method="none") | Statistical Test | https://osf.io/4ya6x/ | R Code First Year Paper Cas Goos Analysis.R |
514 | chi squared test of criteria met by association type | f_statistic <- chisq.test(data_hypotheses_long$association_type, data_hypotheses_long$reporting_criteria_met) f_statistic | Statistical Test | https://osf.io/4ya6x/ | R Code First Year Paper Cas Goos Analysis.R |
515 | Hastags wordcloud of most common hashtags | frequenthashes <- hashtable[hashtable>floor(quantile(hashtable,.999))] hashvec <- unlist(lapply(frequenthashes,function(x) rep(names(x),x))) stretched <- log10(frequenthashes-min(frequenthashes)+1)/max(log10(frequenthashes-min(frequenthashes)+1)) cols <- rgb(1- stretched,0,0) png("hashcloud.png",width = 12,heigh=12,units = "cm",res=300) wordcloud(names(frequenthashes),frequenthashes,random.order = F,colors=cols) dev.off() | Visualization | https://osf.io/u3wby/ | twitter_analyze.R |
516 | Fit the model using unweighted least square (ULS) | m1.fit <- sem( m1, sample.cov = S, sample.nobs = 200, estimator = "ULS") | Statistical Modeling | https://osf.io/vy763/ | Acloserlookatfixed-effectsregressioninsem_supplementarymaterials.R |
517 | Remove items that do not occur in the country | items <- items[items$item %in% colnames(x_country), ] | Data Variable | https://osf.io/8fzns/ | 3H_IRT_helper.R |
518 | multilevel model (repeated measures) correlation using package rmcorr: | td_corr <- rmcorr(subject, trust, distrust, td) td_corr | Statistical Modeling | https://osf.io/kwp6n/ | Everyday_Trust_Rcode.R |
519 | Robust growth model MLR estimator | rctBoot <- growth(rctModel, data=rctWide, se = "bootstrap") summary(rctBoot) rctMLR <- growth(rctModel, data=rctWide, estimator = "MLR") summary(rctMLR) | Statistical Modeling | https://osf.io/fbj3z/ | R Script for Field & Wilcox (2017).R |
520 | create a scatterplot of sleep quality and insomnia scores (FIRST). Add appropriate labels | plot(Sleep_and_COVID19$PSQI_global, Sleep_and_COVID19$FIRST, main = "Main title", xlab = "X axis title", ylab = "Y axis title", frame = FALSE) | Visualization | https://osf.io/94jyp/ | Ex2_BasicAnalysis_Answers.R |
521 | 4) Ttest Examine whether there is a significant difference in sleep quality in those who may have symptoms or been diagnosed with COVID | t.test(PSQI_global ~ SymptomsOrPositive, data = Sleep_and_COVID19) | Statistical Test | https://osf.io/94jyp/ | Ex2_BasicAnalysis_Answers.R |
522 | Examine whether there is a significant difference in mean sleep difficulty scores in those who undertake shiftwork | t.test(PSQI_global ~ ShiftWork, data = Sleep_and_COVID19) ?t.test | Statistical Test | https://osf.io/94jyp/ | Ex2_BasicAnalysis_Answers.R |
523 | 6) Regression Analysis Using multiple regression analysis, examine the influence of the age, hours worked, physical health, coffee drank alertness and insomnia proneness on sleep quality | model1 <- lm(PSQI_global ~ Age + WorkHour + Health + CoffeeToday + StanfordSleepinessScale + FIRST, data = Sleep_and_COVID19) summary(model1) | Statistical Modeling | https://osf.io/94jyp/ | Ex2_BasicAnalysis_Answers.R |
524 | Draw samples setting heart_rate_traj above as mean with appropriate constant variance (10 beats/min) | heart_rate <- rnorm(n = length(heart_rate_traj), mean = heart_rate_traj, sd = 10) | Statistical Modeling | https://osf.io/skp56/ | CSEP_DataScienceExPhys_DataGen_Akerman.R |
525 | Computing pvalues based on the Fdistribution | p_durable <- pf(q=Femp_durable, df1=Df_aest_durable, df2=Df_residual, lower.tail = F) p_nond <- pf(q=Femp_nond, df1=Df_aest_nond, df2=Df_residual, lower.tail = F) | Statistical Test | https://osf.io/n3zfp/ | Aesthetic-Fidelity-Effect-Statistical Analyses.r |
526 | Logistic Regression Model Creating null model and a logistic regression model based on all variables | lr.null <- glm(Treatment ~ 1, data = dat2, family = binomial(link="logit")) lrm <- glm(Treatment ~ Sq + Ssk + Sku + Sp + Sv + Sz + Sa,data = dat2, family = binomial(link="logit")) summary(lrm) | Statistical Modeling | https://osf.io/fvw2k/ | Murrayetal_SurfaceR_Code.R |
527 | Filter and select variables of interest | d <- d_2011 %>% subset(sns == 1) %>% # indicated to use social media subset(!is.na(beh_disclose_soc)) %>% select(id, country, year, gender, age, cons_record, cons_zweck, cons_target, contains("soc"), v143:v156, # beh_setting, int_home, int_work, sns, nth_arg) %>% mutate(cons_mean = (cons_record + cons_zweck + cons_target)/3, cons_mean_bin = ifelse(cons_mean > median(cons_mean, na.rm = T), 1, ifelse(is.na(cons_mean), NA, 0))) %>% mutate(age_group = ifelse(age <= 32, "younger than 32 years", "older than 32 years")) | Data Variable | https://osf.io/m72gb/ | analysis_disclosure.r |
528 | Confirmatory factor analyses CFA for privacy concerns (tauequivalent) | cfa.model <- " priv_con =~ 1*cons_record + 1*cons_zweck + 1*cons_target " fit <- cfa(cfa.model, d_2011) summary(fit, std = T, fit = T) reliability(fit) | Statistical Modeling | https://osf.io/m72gb/ | analysis_disclosure.r |
529 | Estimate models across all specifications Customized model functions to include random country effect | linear <- function(formula, data) { pb$tick() # set tick for progress bar lmer(paste(formula, "+ (1|country)"), data = data) } | Statistical Modeling | https://osf.io/m72gb/ | analysis_disclosure.r |
530 | Decomposing variance Estimate multilevel model (without predictors) | m_lin <- lmer(stdcoef ~ 1 + (1|x) + (1|controls) + (1|subsets) + (1|x:controls) + (1|x:subsets) + (1|subsets:controls), data = results) summary(m_lin) | Statistical Modeling | https://osf.io/m72gb/ | analysis_disclosure.r |
531 | run polynomial regresion as a OLS linear model | f <- paste0(DV, " ~ ", paste("1", IV1, IV2, IV12, IV_IA, IV22, sep=" + "), controlvariables) lm.full <- lm(f, data=df, na.action=na.exclude) | Statistical Modeling | https://osf.io/m6pb2/ | helpers.R |
532 | Matrix of phenotypic change trajectories X and Z, interlineage correlation matrix C, and intertrait crossproduct matrix A, as well as the number of lineages n ( 13) and dimensionality of the vectors p ( 76) | X <- f.mean.diff(data.sc, fishID) Z <- X / sqrt(diag(tcrossprod(X))) C <- tcrossprod(Z) A <- crossprod(Z) n <- nrow(Z) p <- ncol(Z) - 4 # -4 accounting for the loss of d.f. from Procrustes alignment | Data Variable | https://osf.io/6ukwg/ | codes_reanalysis.R |
533 | Histograms comparing dissimilarity between the original and bootstrap replicates between "Moo" and "Pac" (largest and smallest dispersion) This shows that the heteroscedasticity is real (not due to visual distortion) | hist(acos(Z[10, ] %*% Zb[10, , ]), breaks = seq(0, pi / 2, 0.1), col = "#FF000080", main = "Pac (red) versus Moo (blue)", xlab = "Angle between original and bootstrap vectors (radian)") hist(acos(Z[7, ] %*% Zb[7, , ]), add = TRUE, breaks = seq(0, pi / 2, 0.1), col = "#0000FF80") | Visualization | https://osf.io/6ukwg/ | codes_reanalysis.R |
534 | Use kNNdistplot(d.cor.umap, k 3) to determine the value for eps | d.cor.dbscan <- dbscan(d.cor.umap, minPts = 3, eps = 1.8) data.table.large.cluster <- data.table[,d.cor.dbscan$cluster == 1] large.cluster.cor.metric <- as.dist(acos(cor(data.table.large.cluster))) large.umap <- umap( large.cluster.cor.metric, min_dist = 1, n_neighbors = 3 ) large.dbscan <- dbscan(large.umap, minPts = 3, eps = 1) large.cols <- rainbow(length(unique(large.dbscan$cluster)), start = 2/6) write.csv( d.cor.umap, 'd_acos_cor_fixed.csv', row.names = F ) write.csv( d.cor.dbscan$cluster, 'dbscan_clustering_fixed.csv' ) write.csv( large.umap, 'd_acos_cor_fixed_large_cluster.csv', row.names = F ) write.csv( large.dbscan$cluster, 'dbscan_clustering_of_large_cluster_fixed.csv' ) | Visualization | https://osf.io/2qjn5/ | dump_csv.R |
535 | Chisquare test: Difference in preference based on outcome, by condition | chisq.test(matrix(c(19,5,5,17), ncol = 2, byrow = T)) oddsratio.wald.outcome <- oddsratio.wald(matrix(c(19,5,5,17), ncol = 2, byrow = T))$measure[2] # Note: I reorganized the matrix so that r is of the same sign as that of Experiment 2 esc_2x2( grp1yes = 19, grp1no = 5, grp2yes = 5, grp2no = 17, es.type = "or" ) # this confirms Wald's odds ratio | Statistical Test | https://osf.io/qrvje/ | beliefhelp15-220615.R |
536 | Two sample ttest: Difference in proportionate looking based on outcome, by condition | twosamplet.outcome <- t.test(pPositiveO ~ Condition, data = bh2zoom, alternative = "two.sided") twosampled.outcome <- cohen.d(pPositiveO ~ Condition, data = bh2zoom)$estimate | Statistical Test | https://osf.io/qrvje/ | beliefhelp15-220615.R |
537 | Assess how well model captured PWLs ' ' This function calculates several indicators that help to assess how well the model captured captWKLs beyond simple layer structure. ' It get's called by the overarching evaluation script `evaluate_main.R` ' ' So, for each WKL the function computes ' ' * correlation factors for likelihood/distribution/sensitvity versus proportion of unstable grid points ' * the maximum proportion of unstable grid points during the lifetime of the WKL ' * the temporal lag between first forecaster concern (> avy problem) and first time that grid points are unstable in that layer ' different timing instances possible: first time (i) at least one grid point unstable, (ii) the majority of grid points is unstable, (iii) more than 50 % of max proportion unstable ' * the temporal lag between latest forecaster concern (> last day of avy problem) and last time that grid points are unstable in that layer ' need to find rules for becoming dormant and waking up again versus becoming inactive ' ' @param VData object;; don't include both primary/secondary and tertiary gtype_ranks! ' ' @export | assessQualityOfcaptWKL <- function(VData, band, stabilityindex = "p_unstable") { gtype_rank <- unique(VData$vframe$gtype_rank) if (any(c("primary", "secondary") %in% gtype_rank) & "tertiary" %in% gtype_rank) stop("You can only provide either primary and/or secondary, or tertiary gtype_ranks in your VData object to this function!") possible_ranks <- c("tertiary", "secondary", "primary") gtype_rank <- possible_ranks[possible_ranks %in% gtype_rank][1] nrow_max <- length(VData$wkl$wkl_uuid) OUTnames <- c( "wkl_uuid", "wkl_iscrust", "wkldate", "wkltag", "nPDays", "nPDays_anyunstable", "nPDays_median", "nPDays_halfofmax", "nPDays_20", "rho_llhd", "rho_dist", "rho_sens", "p_llhd", "p_dist", "p_sens", "pu_max", "offset_maxes", "offset_mean", "pcapt_max", "lagA_anyunstable", "lagA_median", "lagA_halfofmax", "lagA_20", "lagZ_anyunstable", "lagZ_median", "lagZ_halfofmax", "lagZ_20", "band", "stabilityindex", "gtype_rank" ) OUT <- matrix(nrow = max(1, nrow_max), ncol = length(OUTnames), dimnames = list(seq(max(1, nrow_max)), OUTnames)) for (i in seq_along(VData$wkl$wkl_uuid)) { wuid <- VData$wkl$wkl_uuid[i] wkldate <- as.character(as.Date(VData$wkl$datetag[i])) wkltag <- paste(format(as.Date(VData$wkl$datetag[i]), "%b %d"), substr(gtype_rank, start = 1, stop = 4)) wkl_iscrust <- as.logical(VData$wkl$iscrust[i]) | Statistical Modeling | https://osf.io/w7pjy/ | assessQualityOfcaptWKL.R |
538 | Correlation of life satisfaction and depression across all measurement occasions | d.all %>% select(contains("fsat"), contains("depr"), id) %>% gather(key, value, -id) %>% separate(key, c("time", "variable")) %>% spread(variable, value) %>% select(depr, fsat) %>% zero_order_corr(print = T, digits = 3) %>% select(`1`) %>% slice(2) | Statistical Test | https://osf.io/fdp39/ | analysis.R |
539 | check if latent covariances are equal across groups | est_svc <- cfa(mod_s,quop_use, estimator = "MLR", missing = "FIML", group = "year", group.equal = c("loadings","intercepts","residuals", "lv.variances","lv.covariances"), cluster = "class") | Statistical Modeling | https://osf.io/vphyt/ | Pandemic_Cohorts_vs_Pre_Pandemic_Cohorts.R |
540 | get factor scores from model est_svcm_sl | fs <- lavPredict(est_svcm_sl, method = "bartlett") | Statistical Modeling | https://osf.io/vphyt/ | Pandemic_Cohorts_vs_Pre_Pandemic_Cohorts.R |
541 | Plot for distribution of emotions Arguments: original ddbb, type of emotion, plot title | create.emotion.plot <- function(case_ddbb, emotion="emotion", plot_title=NULL){ | Visualization | https://osf.io/unxj2/ | functions_1.R |
542 | Table and plot for effect sizes Arguments: original ddbb, plot title | create.effectsize.tableplot <- function(case_ddbb, plot_title = NULL){ | Visualization | https://osf.io/unxj2/ | functions_1.R |
543 | Convert average time into a ordinal variable | Tbl2$DurSecAvg1Cat <- cut(Tbl2$DurSecAvg1, breaks = quantile(Tbl2$DurSecAvg1, c(0, 1/3, 2/3, 1)), include.lowest = T) aggregate(Tbl2$DurSecAvg1, list(Tbl2$DurSecAvg1Cat), summary) levels(Tbl2$DurSecAvg1Cat) <- c("Less", "Avg", "More") | Data Variable | https://osf.io/n5j3w/ | 2021-12-03_AnalysisCode.R |
544 | Plotting Figure 3 | png(filename = "Fig03_Demographics.png", width = 19, height = 10, units = "cm", res = 500, pointsize = 7) par(mfrow=c(2,3)) | Visualization | https://osf.io/n5j3w/ | 2021-12-03_AnalysisCode.R |
545 | Predict node for checking differences in distributions with chisquared test | Tbl2$NodeRate <- predict(Engage, newdata = Tbl2, type = "node") table(Tbl2$NodeRate) (temp <- chisq.test(Tbl2$NodeRate, Tbl2$FeedbackEngage)) temp$observed round(100*prop.table(temp$observed, 1), 1) mosaic(temp$observed, shade = T) Tbl2$NodeRate <- predict(FeedTree, newdata = Tbl2, type = "node") table(Tbl2$NodeRate) (temp <- chisq.test(Tbl2$NodeRate, Tbl2$Useful_FeedbackMiddle)) temp$observed round(100*prop.table(temp$observed, 1), 1) mosaic(temp$observed, shade = T) Tbl2$NodeRate <- predict(ExTree, newdata = Tbl2, type = "node") table(Tbl2$NodeRate) (temp <- chisq.test(Tbl2$NodeRate, Tbl2$Useful_FeedbackMiddle)) temp$observed round(100*prop.table(temp$observed, 1), 1) mosaic(temp$observed, shade = T) head(Tbl2$RankTaskPerform) | Statistical Test | https://osf.io/n5j3w/ | 2021-12-03_AnalysisCode.R |
546 | Convert `from_name` to factor with three levels. (Facilitates data visualization). | Campaign_Messages$from_name <- factor(Campaign_Messages$from_name, levels = c ("Stephen Harper", "Justin Trudeau", "Tom Mulcair")) | Data Variable | https://osf.io/3fnjq/ | campaign_messages.R |
547 | change the title text to size 20 and bold | axis.title = element_text(size = 20, face = "bold"), axis.title = element_text(size = 20, face = "bold"), | Visualization | https://osf.io/9e3cu/ | visualization_code.R |
548 | change the axis label text to size 14, bold, and black color | axis.text.x = element_text(size = 14, face = "bold", color = "black"), axis.text.y = element_text(size = 14, face = "bold", color = "black")) + axis.text.x = element_text(size = 14, face = "bold", color = "black"), axis.text.y = element_text(size = 14, face = "bold", color = "black")) + scale_y_continuous(limits = c(0, 17), expand = c(0,0)) + | Visualization | https://osf.io/9e3cu/ | visualization_code.R |
549 | Identify all neighbours within 2km | sp_nb <- spdep::dnearneigh(coords, d1 = 0, d2 = km, row.names = row.names(sp), longlat = TRUE) | Data Variable | https://osf.io/hfjgw/ | 00-utils.r |
550 | take the inverse of the distances | W <- lapply(dsts, function(x) 1 / x) W <- lapply(dsts, function(x) 1 / x) | Data Variable | https://osf.io/hfjgw/ | 00-utils.r |
551 | recode base price and overage as numeric (without $ sign) | x$b1=as.numeric(sapply(strsplit(as.character(x$b1),"\\$"),function(y) y[2])) x$b2=as.numeric(sapply(strsplit(as.character(x$b2),"\\$"),function(y) y[2])) x$b3=as.numeric(sapply(strsplit(as.character(x$b3),"\\$"),function(y) y[2])) x$c1=as.numeric(sapply(strsplit(as.character(x$c1),"\\$"),function(y) y[2])) x$c2=as.numeric(sapply(strsplit(as.character(x$c2),"\\$"),function(y) y[2])) x$c3=as.numeric(sapply(strsplit(as.character(x$c3),"\\$"),function(y) y[2])) | Data Variable | https://osf.io/wbyj7/ | read-data-Exp1.r |
552 | create dataset with possible k values, their prior and likelihood probabilities likelihood is drawn from 1000 random samples from beta distibution | df <- tibble(k = hypo_k) %>% mutate(model_vote = 0.0727^k) %>% mutate(occur = map_int(.$model_vote, ~ length(which(polls$value >= .x))), prior = prob_k) %>% mutate(like = map(.$occur,~ rbeta(trials, .x, polls_len - .x))) | Statistical Modeling | https://osf.io/d4hjq/ | 02_parameter_estimation.R |
553 | visualise normalised posterior on a graph | p <- df %>% ggplot(aes(k, norm_posterior)) + geom_point() + geom_line() p + labs( title = "Probability of k exponent in the interval <0.89;; 0.99>", x = "Value o k exponent", y = "Normalised probability" ) + theme_minimal() | Visualization | https://osf.io/d4hjq/ | 02_parameter_estimation.R |
554 | 3) CALCULATION OF ESTIMATED MARGINAL MEANS Tests for differences in response to jargon for plot | at_jarg <- list(Jargon = c("More", "Less"), UnderN = "Easy", RecognN = "Fairly", TrvlAdv_Atten = "Cons", cut = "3|4") grid_jarg <- ref_grid(clmm_usejar, mode = "exc.prob", at = at_jarg) (emm_jarg <- emmeans(grid_jarg, specs = pairwise ~ Jargon, by = "BackgrAvTraining")) plot_jarg <- summary(emm_jarg$emmeans) cont_jarg <- summary(emm_jarg$contrasts) | Statistical Modeling | https://osf.io/aczx5/ | 220423_Fig04_Use_JargExpl.R |
555 | Gradient of FDR with respect to alpha (and its implied power) across ncp for a onesided ztest Note that the gradient is increasing towards $\infty$ from both side at ncp 0 and alpha 0, but is equal to 0 at that point. The figure does not allow to depict it properly and it seems like as if it peaked at ncp around .5 and was decreasing towards 0. Green corresponds to $P(H_0) .2$, blue to $P(H_0) .5$, and red to $P(H_0) .8$. | alpha <- seq( 0, 1, length.out = 500) ncp <- seq(-5, 5, length.out = 500) FDR.2od <- fill_gradFDR.alpha(.2, alpha, ncp, FALSE) FDR.5od <- fill_gradFDR.alpha(.5, alpha, ncp, FALSE) FDR.8od <- fill_gradFDR.alpha(.8, alpha, ncp, FALSE) plot_ly(x = ~alpha, y = ~ncp, z = ~FDR.5od) %>% add_surface(contours = list( y = list( show = TRUE, project = list(y = TRUE), usecolormap = FALSE, color = "blue" ), x = list( show = TRUE, project = list(x = TRUE), usecolormap = FALSE, color = "blue" ) ), opacity = .4) %>% layout(scene = list( xaxis = list( tickvals = c(0, .05, .1, .2, .5, 1) ), yaxis = list( tickvals = pretty(ncp) ), zaxis = list( tickvals = seq(0, 1, .2), title = list(text = "FDR'") ) )) %>% add_surface(x = ~alpha, y = ~ncp, z = ~FDR.8od, opacity = .4, contours = list( y = list( show = TRUE, project = list(y = TRUE), color = "red" ), x = list( show = TRUE, project = list(x = TRUE), color = "red" ) )) %>% add_surface(x = ~alpha, y = ~ncp, z = ~FDR.2od, opacity = .4, contours = list( y = list( show = TRUE, project = list(y = TRUE), color = "green" ), x = list( show = TRUE, project = list(x = TRUE), color = "green" ) )) | Visualization | https://osf.io/kbjw9/ | 3DFDRPlot.R |
556 | List that will include table (each element should be vector of 3): | Table <- list( c("Network Characteristics",title,""), c("Comparing Global Characteristics",nameOrig, nameRepl) ) | Data Variable | https://osf.io/akywf/ | splithalf_table_function.R |
557 | The row with "def_cond" is moved to a new column and duplicated (this is done with reference to "cond_to_group") | df_tmp_2 <- full_join( df_tmp_1 %>% filter({{cond_to_compare}} != def_cond), df_tmp_1 %>% filter({{cond_to_compare}} == def_cond), by = col ) | Data Variable | https://osf.io/4fvwe/ | return_BF_ttests.R |
558 | Repeating the analyses with closeness as covariate | cor.test(data_3A$closeness_1,data_3A$closeness_2) cor.test(data_3B$closeness_1,data_3B$closeness_2) cor.test(data_3C$closeness_1,data_3C$closeness_2) | Statistical Modeling | https://osf.io/sb3kw/ | Meta_Analysis.R |
559 | check if the number of processed partners is smaller than the number of listed partners as saved in the counter variable | nrow(Relationship_details) < My_EHC_survey$count_SP | Data Variable | https://osf.io/y5gr9/ | Skip_Backwards.R |
560 | run anovas to check interaction between specificity and condition LA | anova_N4_la_sp = summary(aov(Voltage ~ Condition *Specificity , data = mydata_n4_la)) anova_N4_la_sp | Statistical Test | https://osf.io/p7zwr/ | N400.R |
561 | Visualize grid cluststers | mgp.u<-c(2.3,0.6,0.5) if(tifit==T){ tiff(tifname,width=14,height=12.5,units="cm",res=600,compression="lzw") } par(oma=c(0,4,4,4)) layout(matrix(1:9,ncol=3,byrow=T),widths=c(1.28,1,1),heights=c(1,1,1.28)) doFloorC(n=1000,eta=1,char=char,mar1=c(gap,3.5,gap,gap),mgp1=mgp.u,xax=F,yax=T,xl="",col2=col2,seq.s=seq.s,seq.c=seq.c,contour=contour) | Visualization | https://osf.io/pvyhe/ | grid_visualize_3by3.R |
562 | Create the dataset for visualizations Summarize the means (and sd for clusters) | means.cl<-tapply(res$clust,paste(res$n,res$h,res$nu,res$eta),mean,na.rm=T) sd.cl<-tapply(res$clust,paste(res$n,res$h,res$nu,res$eta),sd,na.rm=T) means.dim<-tapply(res$explained3D,paste(res$n,res$h,res$nu,res$eta),mean,na.rm=T) means.avg<-tapply(log(res$avgdist),paste(res$n,res$h,res$nu,res$eta),mean,na.rm=T) vars<-strsplit(names(means.cl)," ") | Data Variable | https://osf.io/pvyhe/ | grid_visualize_3by3.R |
563 | Graph of the distribution of pvalues for each effect | pvalue.plot <- ggplot(melt_pvalue,aes(x = value)) + facet_wrap(~variable, ncol = 5) + geom_histogram() | Visualization | https://osf.io/unxj2/ | functions_2.R |
564 | Graph of the distribution of standard estimates for each effect | est.std.plot <- ggplot(melt_est.std,aes(x = value)) + facet_wrap(~variable, ncol = 5) + geom_histogram() list.result <- list("dist_pvalue.plot" = pvalue.plot, "dist_est.std.plot" = est.std.plot) return(list.result) | Visualization | https://osf.io/unxj2/ | functions_2.R |
565 | returns plot of pvalue disttribution, plot of standard estimates distribution | list.result <- list("dist_pvalue.plot" = pvalue.plot, "dist_est.std.plot" = est.std.plot) return(list.result) | Visualization | https://osf.io/unxj2/ | functions_2.R |
566 | Model estimates plot Arguments: standard estimates distribution, pvalues distributions, use sign level TREU/FALSE, level of signicance, plot title | create.model_estimates.plot <- function (est.std_dist, pvalue_dist, significant = FALSE, sig_level=.05, title = NULL){ | Visualization | https://osf.io/unxj2/ | functions_2.R |
567 | Median of each standard estimates and TRUE/FALSE values if mean(pvalues) < sig_level | est.std_All <- data.frame(cbind(colMeans(est.std_dist), colMeans(pvalue_dist) <= sig_level)) | Statistical Test | https://osf.io/unxj2/ | functions_2.R |
568 | Change the participantID to a shorter & uniform name "Subject" | names(d)[names(d) == "Participant.Public.ID"] <- "Subject" | Data Variable | https://osf.io/cd5r8/ | R_Code_Hui_et_al.R |
569 | For trimmed set, we remove any RT that is too long (>2500ms) | d_trimmed = d %>% filter (acc == 2, Reaction.Time > 300, Reaction.Time < 2500) | Data Variable | https://osf.io/cd5r8/ | R_Code_Hui_et_al.R |
570 | Plotting create a dataset that contains means of both oddnumbered & evennumbered items' RTs | raw_untrimmed<-inner_join(x= mean.o, y=mean.e, by="Subject") names(raw_untrimmed) <- c("Subject", "Odd", "Even") | Visualization | https://osf.io/cd5r8/ | R_Code_Hui_et_al.R |
571 | AVERAGE BY PARTICIPANT | mean.e = de %>% group_by(Subject) %>% summarise(mean(rt.diff)) mean.o = do %>% group_by(Subject) %>% summarise(mean(rt.diff)) mean.e = de %>% group_by(Subject) %>% summarise(mean(zrt.diff)) mean.o = do %>% group_by(Subject) %>% summarise(mean(zrt.diff)) mean.e = de %>% group_by(Subject) %>% summarise(mean(zrt.diff)) mean.o = do %>% group_by(Subject) %>% summarise(mean(zrt.diff)) | Data Variable | https://osf.io/cd5r8/ | R_Code_Hui_et_al.R |
572 | combine them into one data frame | person <- inner_join(person_rand_e, person_rand_o, by = "Subject", suffix = c("_even", "_odd")) | Data Variable | https://osf.io/cd5r8/ | R_Code_Hui_et_al.R |
573 | Wilcoxon ranksum test on widths of unbroken artefacts: | results <- wilcox.test(An2entire$Width,Gmeentire$Width) Z <- qnorm(results$p.value) | Statistical Test | https://osf.io/tgf3q/ | SI_4_Statistical_analyses.R |
574 | Wilcoxon ranksum test on lengths of unbroken artefacts: | results <- wilcox.test(An2entire$Length,Gmeentire$Length) Z <- qnorm(results$p.value) | Statistical Test | https://osf.io/tgf3q/ | SI_4_Statistical_analyses.R |
575 | DATA DOWNSAMPLING add collumns with delta distance and delta time for downsampling | trajectory.df <- trajectory.df %>% group_by(id) %>% mutate(time_diff = difftime(dt, lag(dt, n = 1L), units = "min"), delta_x = x_utm - lag(x_utm, n = 1L), delta_y = y_utm - lag(y_utm, n = 1L), dist = sqrt(delta_x^2+delta_y^2)) | Data Variable | https://osf.io/3bpn6/ | af_homing_dataproc.R |
576 | Caculate daily cumulative distances and time lags Calculate speed Join with relocation info Add consecutive number per individual to indicate tracking day | daily <- trajectory.df %>% group_by(id, date) %>% filter(!dist == "NA") %>% dplyr::summarize(daily_dist = sum(dist), sex = first(sex), time_lag_min = sum(time_lag_min)) %>% mutate(id_day = paste(id, date), daily_speed = (daily_dist/as.numeric(time_lag_min) * 60)) %>% left_join(relocs.perday) %>% group_by(id) %>% mutate(trans_day = row_number()) | Data Variable | https://osf.io/3bpn6/ | af_homing_dataproc.R |
577 | Split up the trajectories into 50m and 200m translocations | trajectory50m <- trajectory.df %>% filter(trans_group == "50m") trajectory200m <- trajectory.df %>% filter(trans_group == "200m") | Data Variable | https://osf.io/3bpn6/ | af_homing_dataproc.R |
578 | Plot daily movement by daily temp & sex | ggplot(data = daily, aes(x = temp, y = daily_dist, group = sex, color = sex)) + geom_point() + stat_smooth(method=lm) + theme_bw() | Visualization | https://osf.io/3bpn6/ | af_homing_dataproc.R |
579 | Plot daily movement by daytime rainfall & sex | ggplot(data = daily, aes(x = rain_daytime, y = daily_dist, group = sex, color = sex)) + geom_point() + stat_smooth(method=lm) + theme_bw() | Visualization | https://osf.io/3bpn6/ | af_homing_dataproc.R |
580 | Plot daily movement by cumulative rainfall & sex | ggplot(data = daily, aes(x = rain_cumul, y = daily_dist, group = sex, color = sex)) + geom_point() + stat_smooth(method=lm) + theme_bw() | Visualization | https://osf.io/3bpn6/ | af_homing_dataproc.R |
581 | x is a matrix containing the data method : correlation method. "pearson"" or "spearman"" is supported removeTriangle : remove upper or lower triangle results : if "html" or "latex" the results will be displayed in html or latex format | corstars <-function(x, method=c("pearson", "spearman"), removeTriangle=c("upper", "lower"), result=c("none", "html", "latex")){ | Statistical Test | https://osf.io/wcfj3/ | funs.R |
582 | Diagonal matrix Tau used to scale the z_w | Tau_w <- matrix(c(DA.mp$`tau_w[1]`,0,0, 0,DA.mp$`tau_w[2]`,0, 0,0,DA.mp$`tau_w[3]` ),nrow=3,ncol=3) | Data Variable | https://osf.io/kdjqz/ | sim_data_LissonEtAl2020.R |
583 | Run fixed effect model | life <- rma(yi = corrs$cor, vi = corrs$vi, measure = "COR", method = "FE") | Statistical Modeling | https://osf.io/9jzfr/ | 20180806funnelplotlifesatisfaction.R |
584 | Set max Y for graph | y_max<-max(d_dist)+1 | Visualization | https://osf.io/ha4q8/ | p_curve_d_distribution_power_app_Lakens.R |
585 | This does the summary. For each group's data frame, return a vector with N, mean, median, and sd | datac <- ddply(data, groupvars, .drop=.drop, .fun = function(xx, col) { c(N = length2(xx[[col]], na.rm=na.rm), mean = mean (xx[[col]], na.rm=na.rm), median = median (xx[[col]], na.rm=na.rm), sd = sd (xx[[col]], na.rm=na.rm) ) }, measurevar ) | Data Variable | https://osf.io/gk6jh/ | summarySE.R |
586 | Rename the "mean" and "median" columns | datac <- rename(datac, c("mean" = paste(measurevar, "_mean", sep = ""))) datac <- rename(datac, c("median" = paste(measurevar, "_median", sep = ""))) datac$se <- datac$sd / sqrt(datac$N) # Calculate standard error of the mean | Data Variable | https://osf.io/gk6jh/ | summarySE.R |
587 | A function that reads in one line in the csv file, add tags and texts around it, and write the output line into an output file | add_tags <- function(l,o){ image_no <- strsplit(as.character(l[1]),"\\.")[[1]][1] image_URL <- paste("<img src='", as.character(l[2]),"'>",sep="") write(paste("[[Block:", image_no, "]] \n [[Question:DB]] \n [[ID:", image_no, "-image]] \n", image_URL, " \n [[Question:TE:SingleLine]] \n [[ID:", image_no, "-TE]] \n", TE, " \n [[Question:MC:SingleAnswer:Vertical]] \n [[ID:", image_no, "-MC]] \n", MC, " \n [[Choices]] \n", choices_formatted, "\n [[Question:Matrix]] \n [[ID:", image_no, "-ratings]] \n ",rating_prompt," \n [[AdvancedChoices]] \n ", rating_statements_formatted, "\n [[AdvancedAnswers]] \n [[Answer]] \n","1 - Strongly Disagree", "\n [[Answer]] \n 2 \n [[Answer]] \n 3 \n [[Answer]] \n 4 \n [[Answer]] \n 5 \n [[Answer]] \n 6 \n [[Answer]] \n", "7 - Strongly Agree"," \n [[PageBreak]] \n", sep = ""), o, append = TRUE) } | Data Variable | https://osf.io/t2jka/ | batchUploadImages.R |
588 | Generate a plot for posterior predictive check (evaluate whether the ' posterior predictive data look more or less similar to the observed data) ' ' @param df dataframe with the data ' @param mod Bayesian model ' ' @return plot with posterior predictive check | get_pp_check <- function(df, mod) { map(list(df), ~brms::pp_check(mod, | Visualization | https://osf.io/5te7n/ | save_get_pp_check.R |
589 | Simulate a single between group study given sample size n, true (fixed) effect size delta, and heterogeneity (random effect) tau. Generate data for the experimental group (y_e) and for the control group (y_c). | y_e = rnorm(n, 0, 1) + delta + rnorm(1, 0, tau) y_c = rnorm(n, 0, 1) | Statistical Modeling | https://osf.io/mg3ny/ | 1_sim_functions.R |
590 | calculate pooled variance S, standardized mean difference d, the variance of d, the pvalue, and N. | S = sqrt(((n - 1) * v_e + (n - 1) * v_c) / df) d = (m_e - m_c) / S var.d = (n + n)/(n * n) + (d^2 / (2 * (n + n))) se.d = sqrt(var.d) N = n + n | Data Variable | https://osf.io/mg3ny/ | 1_sim_functions.R |
591 | Box and Whisker Plots to Compare Models | scales <- list(x=list(relation="free"), y=list(relation="free")) bwplot(results, scales=scales) | Visualization | https://osf.io/uxdwh/ | code.R |
592 | stimuli specify correlations for rnorm_multi (one of several methods) | stim_cors = stim_i_cor stim = rnorm_multi( n = stim_n, vars = 2, r = stim_cors, mu = 0, # means of random intercepts and slopes are always 0 sd = c(stim_sd, stim_version_sd), varnames = c("stim_i", "stim_version_slope") ) %>% mutate( stim_id = 1:stim_n ) stim_cors = stim_i_cor stim = rnorm_multi( n = stim_n, vars = 2, r = stim_cors, mu = 0, # means of random intercepts and slopes are always 0 sd = c(stim_sd, stim_version_sd), varnames = c("stim_i", "stim_version_slope") ) %>% mutate( stim_id = 1:stim_n ) stim_cors = stim_i_cor stim = rnorm_multi( n = stim_n, vars = 2, r = stim_cors, mu = 0, # means of random intercepts and slopes are always 0 sd = c(stim_sd, stim_version_sd), varnames = c("stim_i", "stim_version_slope") ) %>% mutate( stim_id = 1:stim_n ) | Data Variable | https://osf.io/cd5r8/ | Sim_Function.R |
593 | calculate trialspecific effects by adding overall effects and slopes | version_eff = stim_version_eff + stim_version_slope + sub_version_slope, version_eff = stim_version_eff + stim_version_slope + sub_version_slope, version_eff = stim_version_eff + stim_version_slope + sub_version_slope, | Statistical Modeling | https://osf.io/cd5r8/ | Sim_Function.R |
594 | Take the list of csv files from above, and read them all into R (purrr::map function) This will download 100 csv files into a list The reduce function will then bind them all together into a dataframe | large_df <- csv_files %>% purrr::map(function(x) { read.csv(x) }) %>% purrr::reduce(cbind) | Data Variable | https://osf.io/skp56/ | CSEP_DataScienceExPhys_Akerman2021.R |
595 | Re structure this dataframe to turn it to long format (pivot_longer) Keep the X columns in, these are the rowvalues we can use to denote a sample The column names contain the participant (P) and the session number (S) It will display with each 1st sample of each session for a given participant, so need to reorder to get sorted by participant, session, then second sample The names_pattern uses regular expresisons to denote that the names will come from the specified area i.e., after the P, which could contain any values which occur before the underscore, and likewise after the S | large_df <- large_df %>% tidyr::pivot_longer(cols = !contains("X"), names_to = c("participant", "session"), names_pattern = c("P(.*)_S(.*)"), values_to = "heart_rate") %>% dplyr::rename(seconds = "X") %>% dplyr::arrange(participant, session, seconds) %>% dplyr::mutate(participant = as.factor(participant), session = as.factor(session)) | Data Variable | https://osf.io/skp56/ | CSEP_DataScienceExPhys_Akerman2021.R |
596 | pool "exp_buc" columns into one column, so that advocacy type is one variable and create new "experienced" column for each level of advocacy type | data <- data %>% pivot_longer(names_to = "advocacytype", values_to = "experienced", graphic_exp_buc:disprotest_exp_buc) | Data Variable | https://osf.io/3aryn/ | 9Graphingspeciesism_Spanish.R |
597 | Create variables for residue that is 'present' vs 'absent' This is used for the tables/descriptives AND for color/filling the figures | data <- dplyr::mutate(data, oro_zero = if_else(valleculae_severity_rating == 0, "Absent", "Present")) data <- dplyr::mutate(data, hypo_zero = if_else(piriform_sinus_severity_rating == 0,"Absent", "Present")) data <- dplyr::mutate(data, epi_zero = if_else(epiglottis_severity_rating == 0, "Absent", "Present")) data <- dplyr::mutate(data, lv_zero = if_else(laryngeal_vestibule_severity_rating == 0, "Absent", "Present")) data <- dplyr::mutate(data, vf_zero = if_else(vocal_folds_severity_rating == 0, "Absent", "Present")) data <- dplyr::mutate(data, sg_zero = if_else(subglottis_severity_rating == 0, "Absent", "Present")) data <- dplyr::mutate(data, pas_zero = if_else(pas_max < 2, "Absent", "Present")) | Data Variable | https://osf.io/4anzm/ | norms_code.R |
598 | Diagnostic tests to determine appropriate number of factors in EFA parallel analysis (includes scree plot) and Very Simple Structure test optimal number of factors is 6 | fa.parallel(mainData[, c(2:56)], fm = 'ml', fa = 'fa') vss(mainData[, c(2:56)], n = 8, rotate = 'oblimin', fm = 'mle') | Statistical Test | https://osf.io/2j47e/ | Factor analysis.R |
599 | replace missings values (9 and 99) with NA | pilot<- na_if(pilot,99) pilot<- na_if(pilot,-9) | Data Variable | https://osf.io/6579b/ | 03_Supplement.R |
600 | subset of participants with less then 4 missings | pilot<- subset(pilot, pilot$missings_a < 4) pilot_c <- pilot[,c("SJT_kb_00007","SJT_kb_00028","SJT_kb_00058", "SJT_kb_00054","SJT_kb_00072","SJT_kb_00026", "SJT_kb_00060","SJT_kb_00053","SJT_kb_00039", "SJT_kb_00027","SJT_kb_00015","SJT_kb_00038", "SJT_kb_00035","SJT_kb_00055","SJT_kb_00010")] pilot$missings_c <- apply(pilot_w,1,function(x) sum(is.na(x))) pilot<- subset(pilot, pilot$missings_c < 4) pilot_es <- pilot[,c("SJT_kb_00103","SJT_kb_00197","SJT_kb_00136", "SJT_kb_00134","SJT_kb_00113","SJT_kb_00128", "SJT_kb_00117","SJT_kb_00127","SJT_kb_00104", "SJT_kb_00108","SJT_kb_00094","SJT_kb_00143", "SJT_kb_00116","SJT_kb_00202","SJT_kb_00175")] pilot$missings_es <- apply(pilot_es,1,function(x) sum(is.na(x))) pilot<- subset(pilot, pilot$missings_es < 4) | Data Variable | https://osf.io/6579b/ | 03_Supplement.R |
Subsets and Splits