-
Notifications
You must be signed in to change notification settings - Fork 0
/
app.py
298 lines (259 loc) · 9.63 KB
/
app.py
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
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
import os
import numpy as np
import dash
from dash import dcc, html
from dash.dependencies import Input, Output, State
import dash_bootstrap_components as dbc
from components.tabs.maps import tab as tab_map
from components.tabs.topic_map import tab as tab_topics
from components.tabs.heatmaps import tab as tab_heat
from components.header import header
from components.footer import footer
from components.figures.barchart import draw_bar
from components.figures.map import draw_map
from components.figures.heatmap import draw_heatmap
from components.settings import labels, meta_topics, extents, region_groups, url_req, url_asset, url_routes, url_base
from components.data import doc_df, topic_df, df, table_df, dts, dt_sum, geojson, country_shapes, yticks, \
xticks, heat_dfs, m
app = dash.Dash(
__name__,
title='Climate and Health',
external_scripts=[
f'{url_req}{url_asset}assets/js/popper.2.11.7.min.js',
],
external_stylesheets=[
f'{url_req}{url_asset}assets/css/bootstrap.5.2.3.css',
f'{url_req}{url_asset}assets/css/computer-modern-web-font/fonts.css'
],
url_base_pathname=url_base,
requests_pathname_prefix=url_req,
routes_pathname_prefix=url_routes,
assets_url_path=url_asset
)
server = app.server
tabs = dcc.Tabs(className='d-flex justify-content-center',
children=[tab_map, tab_heat, tab_topics])
app.layout = html.Div([
dbc.Container([
header,
dbc.Col(tabs),
], id='mainContainer'),
footer
])
@app.callback(
[
Output('bar', 'figure'),
Output('map', 'figure'),
Output('map-selection', 'data'),
Output('topic-map-selection', 'children'),
Output('map-store', 'data')
],
[
Input('region-select', 'value'),
Input('map', 'selectedData'),
Input('bar', 'clickData'),
Input('rel_slider', 'value'),
Input('barfilter-0', 'n_clicks'),
Input('barfilter-1', 'n_clicks'),
Input('barfilter-2', 'n_clicks'),
Input('barfilter-3', 'n_clicks'),
Input('barfilter-4', 'n_clicks'),
Input('bar', 'relayoutData'),
Input('clear-topics-map-selection', 'n_clicks')
],
[
State('map-store', 'data'),
]
)
def region_interaction(
i, selectedData, clickData, relThreshold,
bf_0, bf_1, bf_2, bf_3, bf_4,
relayoutData, clearTopics, storeData):
rel_ids = doc_df.loc[doc_df['prediction'] >= 1 - relThreshold, 'id']
ctx = dash.callback_context
if 'region-select' in ctx.triggered[0]['prop_id']:
relayoutData = None
selectedData = None
topic_selection = 'No topics selected'
place_ids = []
# did we just, or have we cleared topics without since clicking on them?
if 'clear-topics' in ctx.triggered[0]['prop_id']:
storeData['topics'] = []
storeData['cleared'] = True
# have we just clicked on a topic bar
if 'bar.clickData' in ctx.triggered[0]['prop_id']:
storeData['cleared'] = False
if clickData is not None:
cd = clickData['points'][0]['label']
if cd in storeData['topics']:
storeData['topics'] = [x for x in storeData['topics'] if x != cd]
else:
storeData['topics'] += [cd]
topicids = set([])
for j, bf in enumerate([bf_0, bf_1, bf_2, bf_3, bf_4]):
if bf is None:
continue
elif bf % 2 == 1:
topicids = topicids | set(topic_df.loc[topic_df['Aggregated meta-topic'] == meta_topics[j], 'id'])
if selectedData is not None:
place_ids = [x['id'] for x in selectedData['points'] if 'id' in x]
if len(place_ids) == 1:
place = df.loc[df['place_doc_id'] == place_ids[0]]
place_ids = df.loc[
(df['lat'] == place.lat.values[0]) &
(df['lon'] == place.lon.values[0]),
'place_doc_id'
]
docids = df.loc[df['place_doc_id'].isin(place_ids), 'doc_id']
rel_df = table_df[
(table_df['id'].isin(docids)) &
(table_df['id'].isin(rel_ids))
]
else:
docids = df.loc[df['DFID region'] == labels[i], 'doc_id']
rel_df = table_df[
(table_df['id'].isin(docids)) &
(table_df['id'].isin(rel_ids))
]
sel_df = df
sel_df['subset'] = 0
sel_df.loc[sel_df['id'].isin(rel_df['id']), 'subset'] = 1
gdt = (sel_df[['id', 'subset']]
.merge(dts, left_on='id', right_on='doc_id')
.groupby(['topic_id', 'subset'])['score']
.sum()
.reset_index()
)
# How many documents are there in each topic
tns = (
rel_df[['id']]
.merge(dts, left_on='id', right_on='doc_id')
.query('score>0.01')
.groupby('topic_id')['score']
.count()
.to_frame(name='count')
.sort_values('count')
.reset_index()
)
gdt['share'] = gdt['score'] / gdt.groupby('subset')['score'].transform('sum')
gdt = gdt.merge(dt_sum)
gdt['deviation'] = gdt['share'] / gdt['total_share']
gdt['ldev'] = np.log(gdt['deviation'])
gdt = gdt[gdt['subset'] == 1]
if len(topicids) > 0:
gdt = gdt[gdt.topic_id.isin(topicids)]
cr = (gdt
.merge(topic_df, left_on='topic_id', right_on='id')
.merge(tns, how='left')
.sort_values('ldev')
)
if len(storeData['topics']) > 0:
topics = topic_df[topic_df['short_title'].isin(storeData['topics'])]['id'].values
topic_selection = ', '.join(storeData['topics'])
for top in topics:
rel_df = rel_df[
(rel_df[top] > 0.01)
]
rel_df['topics'] = rel_df[topics].sum()
rel_df = rel_df.sort_values('topics', ascending=False)
if len(place_ids) > 0:
place_ids = set(place_ids) & set(df.loc[df['doc_id'].isin(rel_df['id']), 'place_doc_id'])
else:
place_ids = set(df.loc[df['doc_id'].isin(rel_df['id']), 'place_doc_id'])
cr = cr.reset_index(drop=True)
sbar = cr.loc[cr['short_title'].isin(storeData['topics'])].index
bar = draw_bar(cr, sbar)
if 'barfilter' not in ctx.triggered[0]['prop_id']:
if relayoutData is not None:
if 'autosize' not in relayoutData:
if 'xaxis.range[0]' in relayoutData:
bar.update_xaxes(range=[
relayoutData['xaxis.range[0]'],
relayoutData['xaxis.range[1]']
])
if 'yaxis.range[0]' in relayoutData:
bar.update_yaxes(range=[
relayoutData['yaxis.range[0]'],
relayoutData['yaxis.range[1]']
])
mapFig, mapTitle = draw_map(
region_groups[i], extents[i], geojson,
country_shapes, df[df['id'].isin(rel_ids)], labels[i],
place_ids
)
if rel_df is not None:
rel_df = rel_df.to_dict('records')
return bar, mapFig, rel_df, topic_selection, storeData
@app.callback(
[
Output('heatmap', 'figure'),
Output('heatmap-selection', 'data'),
Output('topic-heatmap-selection', 'children'),
Output('heatmap-store', 'data')
],
[
Input('heatmap-select', 'value'),
Input('heatmap', 'clickData'),
# Input('rel_slider_heatmap', 'value'),
Input('bnorm-1', 'n_clicks'),
Input('bnorm-2', 'n_clicks'),
Input('bnorm-3', 'n_clicks'),
Input('clear-topics-heatmap-selection', 'n_clicks')
# clear topics
],
[
State('heatmap-store', 'data')
]
)
def heatmap_click(
i, clickData,
# relThreshold,
bn1, bn2, bn3, clearTopics, storeData
):
t1, t2, topic_selection, rel_df = None, None, None, None
ctx = dash.callback_context
# rel_ids = doc_df.loc[doc_df['prediction']>=1-relThreshold,'id']
# did we just, or have we cleared topics without since clicking on them?
if 'clear-topics' in ctx.triggered[0]['prop_id']:
clickData = None
storeData['cleared'] = True
if ctx.triggered[0]['prop_id'] != 'heatmap.clickData':
if storeData['cleared']:
clickData = None
else:
storeData['cleared'] = False
for button, bnorm in zip(['bnorm-1', 'bnorm-2', 'bnorm-3'], [-1, 0, 1]):
if button in ctx.triggered[0]['prop_id']:
storeData['bnorm'] = bnorm
# Have we just or have we since clearing, clicked on a topic combination
if clickData is not None:
t1 = clickData['points'][0]['x']
t2 = clickData['points'][0]['y']
topic_selection = f'{t1} & {t2}'
sub_df = heat_dfs[i]
# sub_df = sub_df[sub_df['id'].isin(rel_ids)]
if t1 in sub_df.columns and t2 in sub_df.columns:
thresh = 0.015
sub_df = sub_df[
(sub_df[t1] > thresh) &
(sub_df[t2] > thresh)
]
sub_df['tp'] = sub_df[t1] * sub_df[t2]
rel_df = (
sub_df
.sort_values('tp', ascending=False)
.reset_index(drop=True)
.merge(table_df, left_on='doc_id', right_on='id')
).to_dict('records')
else:
# We must have changed heatmap
t1, t2, topic_selection = None, None, None
heatmap = draw_heatmap(m[i], xticks[i], yticks[i], storeData['bnorm'], t1, t2)
return heatmap, rel_df, topic_selection, storeData
if __name__ == '__main__':
host = os.getenv('CHA_HOST', '127.0.0.1')
port = os.getenv('CHA_PORT', 8058)
debug = os.getenv('CHA_DEBUG', 'off')
app.run_server(debug=debug == 'on',
port=port,
host=host)