-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathhomepage.py
164 lines (136 loc) · 8.3 KB
/
homepage.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
import streamlit as st
from opsci_toolbox.helpers.common import load_pickle, read_json
from opsci_toolbox.helpers.nlp import sample_most_engaging_posts
from opsci_toolbox.helpers.dataviz import subplots_bar_per_day_per_cat, create_scatter_plot, add_shape, pie, network_graph
from opsci_toolbox.helpers.sna import *
from opsci_toolbox.helpers.nlp import load_stopwords_df
from eldar import Query
import plotly.express as px
import pickle
def format_number(number):
if number < 1000:
return str(number)
elif number < 1000000:
return f"{number / 1000:.1f}K"
elif number < 1000000000:
return f"{number / 1000000:.1f}M"
else:
return f"{number / 1000000000:.1f}B"
def main():
st.set_page_config(
page_title="Search engine",
page_icon="🧊",
layout="wide",
initial_sidebar_state="expanded",
)
st.sidebar.title('Settings')
###############################################
# LOAD DATA
###############################################
df = load_pickle("data/df_prod_v2.pickle")
df["datetime"]= pd.to_datetime(df["date"])
plateforme_color_palette = read_json("data/plateforme_color_palette.json")
###############################################
# SIDEBAR SETTINGS / PARAMETERS
###############################################
txt_query_telegram = st.sidebar.text_area("Search on Telegram", value="macron", height=None, max_chars=None, key=None, help=None, on_change=None, args=None, kwargs=None, placeholder=None, disabled=False)
txt_query_twitter = st.sidebar.text_area("Search on Twitter", value="macron", height=None, max_chars=None, key=None, help=None, on_change=None, args=None, kwargs=None, placeholder=None, disabled=False)
date = st.sidebar.date_input("Timerange", value = [df['datetime'].min(), df['datetime'].max()], min_value=df['datetime'].min(), max_value=df['datetime'].max())
lang = st.sidebar.selectbox("Language", ['english', 'russian'], index = 0)
ignore_case = st.sidebar.toggle("Ignore case", value=True)
ignore_accent = st.sidebar.toggle("Ignore accent", value=True)
match_word = st.sidebar.toggle("Match words", value=True)
rolling_period = st.sidebar.text_input("Rolling period", value='7D')
###############################################
# DATA FILTERING
###############################################
if lang == "english":
col_search = "translated_text"
else :
col_search = "text"
df['datetime'] = df['datetime'].dt.date
df = df[(df['datetime'] >= date[0]) & (df['datetime'] <= date[1])]
df_twitter = df[df['plateforme']=="Twitter"].reset_index(drop=True)
df_telegram = df[df['plateforme']=="Telegram"].reset_index(drop=True)
boolean_query_telegram = Query(txt_query_telegram, ignore_case=ignore_case, ignore_accent=ignore_accent, match_word=match_word)
df_telegram = df_telegram[df_telegram[col_search].apply(boolean_query_telegram)]
boolean_query_twitter = Query(txt_query_twitter, ignore_case=ignore_case, ignore_accent=ignore_accent, match_word=match_word)
df_twitter = df_twitter[df_twitter['text'].apply(boolean_query_twitter)]
df_new = pd.concat([df_telegram, df_twitter]).reset_index()
###############################################
# DATA FILTERING
###############################################
total_posts_telegram = df[df['plateforme']=='Telegram']['message_id'].nunique()
total_channels_telegram = df[df['plateforme']=='Telegram']['user_id'].nunique()
sum_views_telegram = df[df['plateforme']=='Telegram']['views'].sum()
sum_eng_telegram = df[df['plateforme']=='Telegram']['engagements'].sum()
total_posts_twitter = df[df['plateforme']=='Twitter']['message_id'].nunique()
total_channels_twitter = df[df['plateforme']=='Twitter']['user_id'].nunique()
sum_views_twitter = df[df['plateforme']=='Twitter']['views'].sum()
sum_eng_twitter = df[df['plateforme']=='Twitter']['engagements'].sum()
################################################
# DATA PREP
################################################
metrics = {
'posts' : ('message_id',"nunique"),
'views': ('views', 'sum'),
'engagements': ('engagements', 'sum'),
'share': ('share', 'sum'),
'likes': ('likes', 'sum'),
'comments': ('comments', 'sum')
}
df_trends_channels = df_new.copy()
df_trends_channels["datetime"]= pd.to_datetime(df_trends_channels["date"])
df_trends_channels.set_index('datetime', inplace=True)
df_trends_channels = df_trends_channels.groupby("plateforme").resample(rolling_period).agg(**metrics).reset_index()
df_trends_channels["datetime"]=df_trends_channels["datetime"].dt.strftime("%Y-%m-%d")
df_trends_channels['color'] = df_trends_channels['plateforme'].map(plateforme_color_palette)
###############################################
# KEY METRICS
###############################################
fig = px.line(df_trends_channels, x='datetime', y='posts', color='plateforme')
st.plotly_chart(fig, use_container_width=True, sharing="streamlit", theme="streamlit")
col1, col2 = st.columns(2, gap="medium")
with col1:
st.title("Telegram")
sub_col1, sub_col2, sub_col3, sub_col4 = st.columns(4, gap="small")
with sub_col1:
st.metric("Verbatims", format_number(df_telegram['message_id'].nunique()), label_visibility="visible")
st.write('{:.2%}'.format(df_telegram['message_id'].nunique()/total_posts_telegram))
with sub_col2:
st.metric("Channels", format_number(df_telegram['user_id'].nunique()), label_visibility="visible")
st.write('{:.2%}'.format(df_telegram['user_id'].nunique()/total_channels_telegram))
with sub_col3:
st.metric("Views", format_number(df_telegram['views'].sum()), label_visibility="visible")
st.write('{:.2%}'.format(df_telegram['views'].sum()/sum_views_telegram))
with sub_col4:
st.metric("Engagements", format_number(df_telegram['engagements'].sum()), label_visibility="visible")
st.write('{:.2%}'.format(df_telegram['engagements'].sum()/sum_eng_telegram))
for i, row in df_telegram.sort_values(by="engagements", ascending=False).iterrows():
st.write('*'*50)
st.write(f"<b>{row['user_name']} - {row['date']}</b>", unsafe_allow_html=True)
st.write(row['translated_text'])
st.write(f"<b>Engagements - {format_number(row['engagements'])} | Views - {format_number(row['views'])} | Shares - {format_number(row['share'])} | Likes - {format_number(row['likes'])} | Comments - {format_number(row['comments'])}</b>", unsafe_allow_html=True)
with col2:
st.title("Twitter")
t_sub_col1, t_sub_col2, t_sub_col3, t_sub_col4 = st.columns(4, gap="small")
with t_sub_col1:
st.metric("Verbatims", format_number(df_twitter['message_id'].nunique()), label_visibility="visible")
st.write('{:.2%}'.format(df_twitter['message_id'].nunique()/total_posts_twitter))
with t_sub_col2:
st.metric("Channels", format_number(df_twitter['user_id'].nunique()), label_visibility="visible")
st.write('{:.2%}'.format(df_twitter['user_id'].nunique()/total_channels_twitter))
with t_sub_col3:
st.metric("Views", format_number(df_twitter['views'].sum()), label_visibility="visible")
st.write('{:.2%}'.format(df_twitter['views'].sum()/sum_views_twitter))
with t_sub_col4:
st.metric("Engagements", format_number(df_twitter['engagements'].sum()), label_visibility="visible")
st.write('{:.2%}'.format(df_twitter['engagements'].sum()/sum_eng_twitter))
for i, row in df_twitter.sort_values(by="engagements", ascending=False).iterrows():
st.write('*'*50)
st.write(f"<b>{row['user_name']} - {row['date']}</b>", unsafe_allow_html=True)
st.write(row['text'])
st.write(f"<b>Engagements - {format_number(row['engagements'])} | Views - {format_number(row['views'])} | Shares - {format_number(row['share'])} | Likes - {format_number(row['likes'])} | Comments - {format_number(row['comments'])}</b>", unsafe_allow_html=True)
st.write(f'<a href="https://www.twitter.com/{row["user_name"]}/status/{row["message_id"]}">Voir le tweet</a>', unsafe_allow_html=True)
if __name__ == "__main__":
main()