首页 文章详情

3个plotly实用进阶范例~

Python与算法之美 | 135 2024-06-05 15:15 0 0 0
UniSMS (合一短信)

本文介绍3个plotly非常实用的高级操作范例:

1,绘制时间序列设置滑块;

2,绘制地图设置高德底图;

3,使用dash构建交互面板;

公众号后台回复关键词:plotly,获取本文jupyter notebook 源代码~

一,绘制时间序列设置滑块

可以使用一个滑块来选择绘图时间范围。

      
      import plotly.express as px 
dfdata = px.data.stocks()

fig = px.line(data_frame=dfdata, x = 'date',y = ['GOOG''AAPL''AMZN''FB''NFLX''MSFT'])
fig.update_xaxes(dtick="M2",tickformat="%Y-%m-%d",rangeslider=dict(visible=True),
             rangeselector={'buttons': [{'count'7,
               'label''1w',
               'step''day',
               'stepmode''backward'},
              {'count'1'label''1m''step''month''stepmode''backward'},
              {'count'6'label''6m''step''month''stepmode''backward'},
              {'count'1'label''1y''step''year''stepmode''backward'},
              {'step''all'}]}
            )
fig.update_layout(#autosize=True,
    #width=1000, 
    #height=600,
    margin=dict(
        r=0, t=0, l=0, b=0, pad=0)
    )
fig.show()
fig.write_html('test.html')

效果如下:

0e082192748aecd016cda989e24b874d.webp

二,绘制地图设置高德底图

plotly绘制地图可以使用高德底图。

      
      
import plotly.express as px 
dfdata = pd.DataFrame({'lat'39 + np.random.rand(100),
                        'lon'116+np.random.rand(100),
                       'color'10*np.random.rand(100),
                       'size'0.5*np.random.rand(100),
                      })

fig = px.scatter_mapbox(dfdata, lat="lat"
                        lon="lon", color="color",
                        size="size",            
                        color_continuous_scale=px.colors.cyclical.IceFire, 
                        size_max=15, zoom=10
                       )
basemap_layer = [
    dict(
        below="traces",
        sourcetype="raster",
        sourceattribution="高德地图",
        source=[
            "http://wprd01.is.autonavi.com/appmaptile?x={x}&y={y}&z={z}&lang=zh_cn&size=1&scl=1&style=7"
        ]
    )
]

fig.update_mapboxes(style='white-bg',zoom=7,layers=basemap_layer)
fig.update_layout(margin=dict(r=0, t=0, l=0, b=0, pad=0))
fig.show()

效果如下:

139ed7caaaafeed03c71aa597e244445.webp

三,使用dash构建交互面板

使用plotly的dash可以让做出非常丰富的前端交互效果。

详情参考:https://dash.plotly.com/

      
      import dash
from dash import Dash, dcc, html, Input, Output
import plotly.express as px


external_stylesheets = ['https://codepen.io/chriddyp/pen/bWLwgP.css']

template_list = ['plotly','ggplot2''seaborn''simple_white',
         'plotly_white''plotly_dark''presentation''xgridoff',
         'ygridoff''gridon']

# 1,生成示例数据
dfdata = px.data.stocks()


# 2, 创建Dash app
app = Dash(__name__,external_stylesheets=external_stylesheets)

# 3, 设计页面布局
app.layout = html.Div([
    html.H3(children='头部互联网美股走势数据'),
    dcc.Graph(id='stock-plot'),
    html.Br(),
    html.Label('template'),
    dcc.Slider(
        id='template',
        min=0,
        max=len(template_list)-1,
        value=0,
        marks={i: template_list[i] for i in range(len(template_list))},
        step=1
    ),
    html.Label('font_size'),
    dcc.Slider(
        id='font_size',
        min=10,
        max=20,
        value=15,
        marks={i: str(i) for i in range(10,21)},
        step=1
    )
    
])

# 4, 编写回调函数
@app.callback(
    Output(component_id='stock-plot', component_property='figure'),
    [Input(component_id='template',  component_property='value'),
     Input(component_id='font_size', component_property='value')
    ]
)
def update_figure(template,font_size):
    fig = px.line(data_frame=dfdata, x = 'date',y = ['GOOG''AAPL''AMZN''FB''NFLX''MSFT'])
    fig.update_xaxes(dtick="M1",tickformat="%Y-%m-%d",rangeslider=dict(visible=True),
                 rangeselector={'buttons': [{'count'7,
                   'label''1w',
                   'step''day',
                   'stepmode''backward'},
                  {'count'1'label''1m''step''month''stepmode''backward'},
                  {'count'6'label''6m''step''month''stepmode''backward'},
                  {'count'1'label''1y''step''year''stepmode''backward'},
                  {'step''all'}]}
                )
    fig.layout.template = template_list[template]
    fig.update_layout(autosize=True,
        #width=1000, 
        #height=600,
        margin=dict(
            r=0, t=0, l=0, b=0, pad=0)
        )
    fig.update_layout({"font.size":font_size})
    return fig


# 5, 运行交互页面
if __name__ == '__main__':
    app.run_server(debug=True)
    #app.run(jupyter_mode="tab") #'inline', 'external', 'jupyterlab', 'tab'

运行上述代码,会弹出一个可以交互的网页,效果如下:

b2d10ce91f5d846ece4e7a98f1074098.webp


猜你喜欢:


Plotly深入浅出


8a344bb99519f16779b59222a73ac73c.webp

good-icon 0
favorite-icon 0
收藏
回复数量: 0
    暂无评论~~
    Ctrl+Enter