File size: 5,827 Bytes
e549c90
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
# AUTOGENERATED! DO NOT EDIT! File to edit: ../../nbs/book/EventTracker/11_Andon.ipynb.

# %% auto 0
__all__ = ['df', 'farms', 'selected_farms', 'filter_df', 'generate_discrete_colors', 'get_discrete_colormap', 'get_summary',
           'get_event_metrics', 'get_algo_performance', 'get_pie', 'get_data', 'get_farms', 'kpi']

# %% ../../nbs/book/EventTracker/11_Andon.ipynb 2
import streamlit as st
import pandas as pd
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import plotly.express as px

# %% ../../nbs/book/EventTracker/11_Andon.ipynb 6
st.set_page_config(
    page_title="Andon",
    layout='wide'
)

# %% ../../nbs/book/EventTracker/11_Andon.ipynb 11
def generate_discrete_colors(n):
    colors = px.colors.qualitative.Plotly
    return [colors[i % len(colors)] for i in range(n)]

generate_discrete_colors(5)

# %% ../../nbs/book/EventTracker/11_Andon.ipynb 12
def get_discrete_colormap(classes):
    colors = generate_discrete_colors(len(classes))
    color_discrete_map = dict(zip(classes, colors))
    return color_discrete_map

# %% ../../nbs/book/EventTracker/11_Andon.ipynb 13
def get_summary(df, col_a='Type of Alert', col_b='Ground Truth'):
    cal_df = df[[col_a, col_b]].groupby([col_a, col_b]).value_counts().reset_index()

    
    color_discrete_map = get_discrete_colormap(df[col_b].dropna().unique())
    fig = px.bar(cal_df, y=col_a, 
                x='count', 
                color=col_b, 
                color_discrete_map=color_discrete_map, 
                orientation='h', 
                title=f'Summary: {col_a}|{col_b}')

    fig.update_layout(
        margin=dict(l=10, r=10, t=40, b=40),
        yaxis=dict(automargin=True)
    )

    return fig

# %% ../../nbs/book/EventTracker/11_Andon.ipynb 16
def get_event_metrics(validated_df, 
                      event_type_cols=['Heat', 'Missed Heat'], 
                      event_type='Heat'):
    val_event_df = validated_df[validated_df['Type of Alert'].isin(event_type_cols)]
    a = val_event_df['Ground Truth'].value_counts()
    
    recall  = a.get('TRUE', 0)*100/(a.get('TRUE', 0)+a.get('MISSED', np.finfo(float).eps))
    precision = a.get('TRUE', 0)*100/(a.get('TRUE', 0)+a.get('FALSE', np.finfo(float).eps))

    return {
        f'Validated_{event_type}_Recall': recall,
        f'Validated_{event_type}_Precision': precision,
    }

# %% ../../nbs/book/EventTracker/11_Andon.ipynb 17
def get_algo_performance(df):
    confirmed_df = df[df['Ground Truth'].isin(['TRUE','FALSE', 'MISSED', 'EXCLUDE', 'OTHER'])]
    validated_df = df[df['Ground Truth'].isin(['TRUE','FALSE', 'MISSED'])]
    missed_df = df[df['Ground Truth'].isin(['MISSED'])]
    heat_metrics = get_event_metrics(validated_df, 
                        event_type_cols=['Heat', 'Missed Heat'], 
                        event_type='Heat')
    health_metrics = get_event_metrics(validated_df, 
                        event_type_cols=['Health', 'Missed Health'], 
                        event_type='Health')

    data_coverage_metrics = {'Total':len(df),
                             'Confirmed': len(confirmed_df),
                             'Validated': len(validated_df),
                             'Missed': len(missed_df),
                             'Confirmed_Percentage': len(confirmed_df)*100/len(df),
                             'Validated_Percentage': len(validated_df)*100/len(df),
                             'Val_On_Confirm_Percentage': len(validated_df)*100/len(confirmed_df),
                             'Missed_On_Val_Percentage': len(missed_df)*100/len(validated_df) }
    performance_metrics = {**data_coverage_metrics,**heat_metrics, **health_metrics}
    return performance_metrics

# %% ../../nbs/book/EventTracker/11_Andon.ipynb 20
def get_pie(df, colname):
    df_dist = df[colname].value_counts().reset_index()
    fig = px.pie(df_dist, values='count', 
                names=colname, 
                title=f'{colname}',
                color = colname,
                color_discrete_map = get_discrete_colormap(df[colname].dropna().unique())
                )
    return fig

# %% ../../nbs/book/EventTracker/11_Andon.ipynb 35
def get_data(url="https://docs.google.com/spreadsheets/d/1TbqmmSzXtY8DwolVo9rN7M1QdZxLG30nro2cqrvn2CA/export?format=csv&#gid=294763682"):
    df = pd.read_csv(url).dropna(how='all')
    df['Alert Date'] = df['Alert Date'].str.replace("//", "-")
    df = df.dropna(subset=['Alert Date'])
    df['Type of Alert'] = df['Type of Alert'].str.strip()
    return df
# colors

# %% ../../nbs/book/EventTracker/11_Andon.ipynb 36
def get_farms(df):
    return df['Dairy Farm'].dropna().unique().tolist()

# %% ../../nbs/book/EventTracker/11_Andon.ipynb 39
def kpi(df, ncols = 7):
    metrics = get_algo_performance(df)
    dcols = st.columns(ncols)
    for (i, (k, v)) in enumerate(metrics.items()):
        with dcols[i%ncols]:
            st.metric(k, np.round(v, 2))

# %% ../../nbs/book/EventTracker/11_Andon.ipynb 40
#|eval: false
df = get_data()
farms = get_farms(df)
selected_farms = st.sidebar.multiselect('Select farms', farms, max_selections=len(farms))
filter_df = df
if selected_farms: 
    str_farms = "|".join(selected_farms)
    st.write(f"Selected_farms: {str_farms}")
    filter_df = df[df['Dairy Farm'].isin(selected_farms)]
else:
    st.write("Selected_farms: All")
# st.write(farms)
kpi(filter_df,ncols=4)
st.plotly_chart(get_pie(filter_df, colname = 'Type of Alert'))
st.plotly_chart(get_pie(filter_df, colname = 'Ground Truth'))
st.plotly_chart(get_pie(filter_df, colname = 'Dairy Farm'))
st.plotly_chart(get_summary(filter_df, col_a='Type of Alert', col_b='Ground Truth'))
st.plotly_chart(get_summary(filter_df, col_b='Type of Alert', col_a='Dairy Farm'))
st.plotly_chart(get_summary(filter_df, col_b='Ground Truth', col_a='Dairy Farm'))
#