233 lines
8.6 KiB
Python
233 lines
8.6 KiB
Python
|
import dash
|
|||
|
|
from dash import dcc, html
|
|||
|
|
from dash.dependencies import Input, Output
|
|||
|
|
import plotly.graph_objs as go
|
|||
|
|
import pandas as pd
|
|||
|
|
import pytz
|
|||
|
|
from datetime import datetime
|
|||
|
|
from sqlalchemy import create_engine
|
|||
|
|
|
|||
|
|
# 数据库连接配置
|
|||
|
|
DB_CONFIG = {
|
|||
|
|
'host': '8.155.23.172',
|
|||
|
|
'port': 3306,
|
|||
|
|
'user': 'root2',
|
|||
|
|
'password': 'tG0f6PVYh18le41BCb',
|
|||
|
|
'database': 'elonX'
|
|||
|
|
}
|
|||
|
|
TABLE_NAME = 'elon_tweets'
|
|||
|
|
|
|||
|
|
# 使用SQLAlchemy创建数据库连接
|
|||
|
|
db_uri = f"mysql+pymysql://{DB_CONFIG['user']}:{DB_CONFIG['password']}@{DB_CONFIG['host']}:{DB_CONFIG['port']}/{DB_CONFIG['database']}"
|
|||
|
|
engine = create_engine(db_uri)
|
|||
|
|
|
|||
|
|
# 加载数据
|
|||
|
|
df = pd.read_sql(f'SELECT timestamp FROM {TABLE_NAME}', con=engine)
|
|||
|
|
|
|||
|
|
# 数据预处理（基于EST）
|
|||
|
|
eastern = pytz.timezone('America/New_York') # EST
|
|||
|
|
pacific = pytz.timezone('America/Los_Angeles') # PST
|
|||
|
|
central = pytz.timezone('America/Chicago') # CST
|
|||
|
|
df['datetime'] = pd.to_datetime(df['timestamp'], unit='s')
|
|||
|
|
df['datetime_est'] = df['datetime'].dt.tz_localize('UTC').dt.tz_convert(eastern)
|
|||
|
|
df['date'] = df['datetime_est'].dt.date
|
|||
|
|
df['minute_of_day'] = df['datetime_est'].dt.hour * 60 + df['datetime_est'].dt.minute
|
|||
|
|
agg_df = df.groupby(['date', 'minute_of_day']).size().reset_index(name='tweet_count')
|
|||
|
|
|
|||
|
|
# 获取所有日期用于选择器
|
|||
|
|
all_dates = sorted(agg_df['date'].unique())
|
|||
|
|
default_dates = all_dates[-4:] # 默认显示最近4天
|
|||
|
|
|
|||
|
|
# 初始化Dash应用
|
|||
|
|
app = dash.Dash(__name__)
|
|||
|
|
|
|||
|
|
# 时间间隔选项
|
|||
|
|
interval_options = [
|
|||
|
|
{'label': '1分钟', 'value': 1},
|
|||
|
|
{'label': '5分钟', 'value': 5},
|
|||
|
|
{'label': '10分钟', 'value': 10},
|
|||
|
|
{'label': '30分钟', 'value': 30},
|
|||
|
|
{'label': '60分钟', 'value': 60}
|
|||
|
|
]
|
|||
|
|
|
|||
|
|
# Dash应用布局
|
|||
|
|
app.layout = html.Div([
|
|||
|
|
html.H1("Elon Musk 发帖时间分析 (EST)"),
|
|||
|
|
dcc.Tabs(id='tabs', value='daily-view', children=[
|
|||
|
|
# 选项卡1：每日视图（折线图）
|
|||
|
|
dcc.Tab(label='Daily View', value='daily-view', children=[
|
|||
|
|
dcc.DatePickerSingle(
|
|||
|
|
id='date-picker',
|
|||
|
|
min_date_allowed=min(all_dates),
|
|||
|
|
max_date_allowed=max(all_dates),
|
|||
|
|
initial_visible_month=max(all_dates),
|
|||
|
|
date=max(all_dates)
|
|||
|
|
),
|
|||
|
|
dcc.Dropdown(
|
|||
|
|
id='daily-interval-picker',
|
|||
|
|
options=interval_options,
|
|||
|
|
value=10, # 默认10分钟
|
|||
|
|
style={'width': '50%'}
|
|||
|
|
),
|
|||
|
|
html.Div(id='daily-tweet-summary', style={'fontSize': 20, 'margin': '10px'}), # 单日汇总
|
|||
|
|
dcc.Graph(id='daily-tweet-graph')
|
|||
|
|
]),
|
|||
|
|
# 选项卡2：多日视图（多线折线图）
|
|||
|
|
dcc.Tab(label='Multi-Day View', value='multi-day-view', children=[
|
|||
|
|
dcc.Checklist(
|
|||
|
|
id='multi-date-picker',
|
|||
|
|
options=[{'label': str(date), 'value': str(date)} for date in all_dates],
|
|||
|
|
value=[str(date) for date in default_dates],
|
|||
|
|
style={'height': '200px', 'overflow': 'auto'}
|
|||
|
|
),
|
|||
|
|
dcc.Dropdown(
|
|||
|
|
id='multi-interval-picker',
|
|||
|
|
options=interval_options,
|
|||
|
|
value=10, # 默认10分钟
|
|||
|
|
style={'width': '50%'}
|
|||
|
|
),
|
|||
|
|
html.Div(id='multi-day-warning', style={'color': 'red'}),
|
|||
|
|
html.Div(id='multi-tweet-summary', style={'fontSize': 20, 'margin': '10px'}), # 多日汇总
|
|||
|
|
dcc.Graph(id='multi-tweet-graph')
|
|||
|
|
])
|
|||
|
|
])
|
|||
|
|
])
|
|||
|
|
|
|||
|
|
|
|||
|
|
# 数据聚合函数，按指定时间间隔分组并填充0
|
|||
|
|
def aggregate_data(data, interval):
|
|||
|
|
all_minutes = pd.DataFrame({'interval_group': range(0, 1440, interval)})
|
|||
|
|
result = []
|
|||
|
|
for date in data['date'].unique():
|
|||
|
|
day_data = data[data['date'] == date].copy()
|
|||
|
|
day_data['interval_group'] = (day_data['minute_of_day'] // interval) * interval
|
|||
|
|
agg = day_data.groupby('interval_group').size().reset_index(name='tweet_count')
|
|||
|
|
complete_data = all_minutes.merge(agg, on='interval_group', how='left').fillna({'tweet_count': 0})
|
|||
|
|
complete_data['date'] = date
|
|||
|
|
result.append(complete_data)
|
|||
|
|
return pd.concat(result, ignore_index=True)
|
|||
|
|
|
|||
|
|
|
|||
|
|
# 生成X轴刻度（EST时间）
|
|||
|
|
def generate_xticks(interval):
|
|||
|
|
ticks = list(range(0, 1440, interval))
|
|||
|
|
tick_labels = [f"{m // 60:02d}:{m % 60:02d}" for m in ticks]
|
|||
|
|
return ticks, tick_labels
|
|||
|
|
|
|||
|
|
|
|||
|
|
# 回调函数1：更新Daily View图表和汇总
|
|||
|
|
@app.callback(
|
|||
|
|
[Output('daily-tweet-graph', 'figure'),
|
|||
|
|
Output('daily-tweet-summary', 'children')],
|
|||
|
|
[Input('date-picker', 'date'),
|
|||
|
|
Input('daily-interval-picker', 'value'),
|
|||
|
|
Input('tabs', 'value')]
|
|||
|
|
)
|
|||
|
|
def update_daily_graph(selected_date, interval, tab):
|
|||
|
|
if tab != 'daily-view':
|
|||
|
|
return go.Figure(), ""
|
|||
|
|
|
|||
|
|
if isinstance(selected_date, str):
|
|||
|
|
selected_date = datetime.strptime(selected_date, '%Y-%m-%d').date()
|
|||
|
|
|
|||
|
|
day_data = agg_df[agg_df['date'] == selected_date].copy()
|
|||
|
|
if day_data.empty:
|
|||
|
|
day_data = pd.DataFrame({'date': [selected_date], 'minute_of_day': [0]})
|
|||
|
|
tweet_count_total = 0
|
|||
|
|
else:
|
|||
|
|
tweet_count_total = day_data['tweet_count'].sum()
|
|||
|
|
|
|||
|
|
agg_data = aggregate_data(day_data, interval)
|
|||
|
|
xticks, xtick_labels = generate_xticks(interval if interval >= 30 else 60)
|
|||
|
|
|
|||
|
|
fig = go.Figure()
|
|||
|
|
fig.add_trace(go.Scatter(
|
|||
|
|
x=agg_data['interval_group'],
|
|||
|
|
y=agg_data['tweet_count'],
|
|||
|
|
mode='lines',
|
|||
|
|
name='推文数量',
|
|||
|
|
line=dict(color='blue')
|
|||
|
|
))
|
|||
|
|
|
|||
|
|
# 计算凌晨2点位置（基于EST）
|
|||
|
|
eastern_2am = eastern.localize(datetime.combine(selected_date, datetime.time(2, 0)))
|
|||
|
|
pacific_2am = pacific.localize(datetime.combine(selected_date, datetime.time(2, 0))).astimezone(eastern)
|
|||
|
|
central_2am = central.localize(datetime.combine(selected_date, datetime.time(2, 0))).astimezone(eastern)
|
|||
|
|
eastern_2am_minute = eastern_2am.hour * 60 + eastern_2am.minute # 120分钟
|
|||
|
|
pacific_2am_minute = pacific_2am.hour * 60 + pacific_2am.minute # 300分钟 (5:00 EST)
|
|||
|
|
central_2am_minute = central_2am.hour * 60 + central_2am.minute # 180分钟 (3:00 EST)
|
|||
|
|
|
|||
|
|
# 添加垂直线
|
|||
|
|
fig.add_vline(x=eastern_2am_minute, line_dash="dash", line_color="red", annotation_text="纽约 2AM")
|
|||
|
|
fig.add_vline(x=pacific_2am_minute, line_dash="dash", line_color="blue", annotation_text="加州 2AM")
|
|||
|
|
fig.add_vline(x=central_2am_minute, line_dash="dash", line_color="green", annotation_text="新奥尔良 2AM")
|
|||
|
|
|
|||
|
|
fig.update_layout(
|
|||
|
|
title=f'{selected_date} 的推文频率（间隔 {interval} 分钟，EST）',
|
|||
|
|
xaxis_title='东部时间 (HH:MM)',
|
|||
|
|
yaxis_title='推文数量',
|
|||
|
|
xaxis=dict(range=[0, 1440], tickvals=xticks, ticktext=xtick_labels, tickangle=45),
|
|||
|
|
height=600
|
|||
|
|
)
|
|||
|
|
|
|||
|
|
summary = f"单日推文总数: {int(tweet_count_total)}"
|
|||
|
|
return fig, summary
|
|||
|
|
|
|||
|
|
|
|||
|
|
# 回调函数2：更新Multi-Day View图表、警告信息和汇总
|
|||
|
|
@app.callback(
|
|||
|
|
[Output('multi-tweet-graph', 'figure'),
|
|||
|
|
Output('multi-day-warning', 'children'),
|
|||
|
|
Output('multi-tweet-summary', 'children')],
|
|||
|
|
[Input('multi-date-picker', 'value'),
|
|||
|
|
Input('multi-interval-picker', 'value'),
|
|||
|
|
Input('tabs', 'value')]
|
|||
|
|
)
|
|||
|
|
def update_multi_graph(selected_dates, interval, tab):
|
|||
|
|
if tab != 'multi-day-view':
|
|||
|
|
return go.Figure(), "", ""
|
|||
|
|
|
|||
|
|
if len(selected_dates) > 10:
|
|||
|
|
selected_dates = selected_dates[:10]
|
|||
|
|
warning = "最多只能选择10天，已自动截取前10天。"
|
|||
|
|
else:
|
|||
|
|
warning = ""
|
|||
|
|
|
|||
|
|
selected_dates = [datetime.strptime(date, '%Y-%m-%d').date() for date in selected_dates]
|
|||
|
|
multi_data = agg_df[agg_df['date'].isin(selected_dates)].copy()
|
|||
|
|
if multi_data.empty:
|
|||
|
|
multi_data = pd.DataFrame({'date': selected_dates, 'minute_of_day': [0] * len(selected_dates)})
|
|||
|
|
tweet_count_total = 0
|
|||
|
|
else:
|
|||
|
|
tweet_count_total = multi_data['tweet_count'].sum()
|
|||
|
|
|
|||
|
|
agg_data = aggregate_data(multi_data, interval)
|
|||
|
|
xticks, xtick_labels = generate_xticks(interval if interval >= 30 else 60)
|
|||
|
|
|
|||
|
|
fig = go.Figure()
|
|||
|
|
for i, date in enumerate(selected_dates):
|
|||
|
|
day_data = agg_data[agg_data['date'] == date]
|
|||
|
|
fig.add_trace(go.Scatter(
|
|||
|
|
x=day_data['interval_group'],
|
|||
|
|
y=day_data['tweet_count'],
|
|||
|
|
mode='lines',
|
|||
|
|
name=str(date),
|
|||
|
|
visible=True if i < 4 else 'legendonly'
|
|||
|
|
))
|
|||
|
|
|
|||
|
|
fig.update_layout(
|
|||
|
|
title=f'多日推文频率对比（间隔 {interval} 分钟，EST）',
|
|||
|
|
xaxis_title='东部时间 (HH:MM)',
|
|||
|
|
yaxis_title='推文数量',
|
|||
|
|
xaxis=dict(range=[0, 1440], tickvals=xticks, ticktext=xtick_labels, tickangle=45),
|
|||
|
|
height=600,
|
|||
|
|
showlegend=True
|
|||
|
|
)
|
|||
|
|
|
|||
|
|
summary = f"所选日期推文总数: {int(tweet_count_total)}"
|
|||
|
|
return fig, warning, summary
|
|||
|
|
|
|||
|
|
|
|||
|
|
# 运行应用
|
|||
|
|
if __name__ == '__main__':
|
|||
|
|
app.run_server(debug=True)
|