悬停情节

Question

我正在绘制 3D 散点图：

d = {'x':[1,2,3,4], 'y':[2,3,1,5], 'z':[3,2,3,2], 't':[4,1,2,3], 'score':[2,3,1,2]}
df = pd.DataFrame (d)

xtitle = 'x'
ytitle = 'y'
ztitle = 'z'

trace1 = go.Scatter3d(x=df[xtitle], 
                        y= df[ytitle], 
                          z = df[ztitle],
                                       marker=dict(color=df['score'],
                                                   showscale=True,
                                                  colorbar=dict(
                                                    title='score)'
                                                )),                       
                                       mode='markers')

layout = go.Layout (
        scene = Scene(
            xaxis = dict (title = xtitle),
            yaxis = dict (title = ytitle),
            zaxis = dict (title = ztitle)
        )
    )
fig = go.Figure(data=[trace1], layout = layout)
plotly.offline.iplot(fig)

当我将鼠标悬停在一个点上时，它会显示 x、y 和 z 值。

在数据框中，df我有另一列名为t，我希望当我将鼠标悬停在一个点上时，它也会显示 x、y、z、t 和分数。

我怎样才能做到这一点？

Answer 1

您可以在跟踪中使用text或hoverinfo。

import plotly.graph_objs as go
from plotly.offline import download_plotlyjs, init_notebook_mode, plot, iplot
init_notebook_mode(connected=True)
import pandas as pd
d = {'x':[1,2,3,4], 'y':[2,3,1,5], 'z':[3,2,3,2], 't':[4,1,2,3], 'score':[2,3,1,2]}
df = pd.DataFrame (d)

xtitle = 'x'
ytitle = 'y'
ztitle = 'z'

trace1 = go.Scatter3d(
    x=df[xtitle], 
    y= df[ytitle], 
    z = df[ztitle],
    marker=dict(
        color=df['score'],
        showscale=True,
        colorbar=dict(title='score)')
    ),
    mode='markers',
    text = ["t: {}".format(x) for x in df['t'] ]  # <-- added line!
    # hoverinfo = df['t']  # alternative
)

layout = go.Layout (
        scene = dict(
            xaxis = dict (title = xtitle),
            yaxis = dict (title = ytitle),
            zaxis = dict (title = ztitle)
        )
    )
fig = go.Figure(data=[trace1], layout = layout)
iplot(fig)

有用的链接：SO answer 0，SO answer 1，SO answer 2。您要使用该text属性，hoverinfo并且hovertext. 这是文档。