python - 如何使用 plotly figure_factory 仅绘制散点图矩阵的下半部分？

Question

我从 ploty figure_factory 看到这个例子

import plotly.graph_objects as go
import plotly.figure_factory as ff

import numpy as np
import pandas as pd

df = pd.DataFrame(np.random.randn(20, 4),
                columns=['Column A', 'Column B', 'Column C', 'Column D'])

df['Fruit'] = pd.Series(['apple', 'apple', 'grape', 'apple', 'apple',
                                'grape', 'pear', 'pear', 'apple', 'pear',
                                'apple', 'apple', 'grape', 'apple', 'apple',
                                'grape', 'pear', 'pear', 'apple', 'pear'])


fig = ff.create_scatterplotmatrix(df, diag='histogram',index='Fruit',
                                  height=800, width=800)
fig.show()

还有剧情:

但我想知道是否可以只绘制下半部分，因为上半部分是相同的。这可以通过 lotly.graph_objects(感谢 showupperhalf=False :例如

import plotly.graph_objects as go
import pandas as pd

df = pd.read_csv('https://raw.githubusercontent.com/plotly/datasets/master/iris-data.csv')
index_vals = df['class'].astype('category').cat.codes

fig = go.Figure(data=go.Splom(
                dimensions=[dict(label='sepal length',
                                 values=df['sepal length']),
                            dict(label='sepal width',
                                 values=df['sepal width']),
                            dict(label='petal length',
                                 values=df['petal length']),
                            dict(label='petal width',
                                 values=df['petal width'])],
                showupperhalf=False, # remove plots on diagonal
                text=df['class'],
                marker=dict(color=index_vals,
                            showscale=False, # colors encode categorical variables
                            line_color='white', line_width=0.5)
                ))


fig.update_layout(
    title='Iris Data set',
    width=600,
    height=600,
)

fig.show()

结果是:

可以和figure_factory一样吗？提前致谢

或者可以将散点图从 go.Splom 更改为直方图？或方框？

Answer 1

简答:

您可以通过将子图中的所有迹线替换为空的迹线对象(例如 go.Scatter()

)来清空您想要的任何子图

---

详情:

Scatterplot Matrix似乎没有对角线上直方图的选项。和 ff.create_scatterplotmatrix似乎没有隐藏部分矩阵的选项。但这并不意味着您在这里尝试做的事情是不可能的。

下面完整的代码片段使用了一个矩阵...

shows = array([[1., 0., 0., 0.],
              [1., 1., 0., 0.],
              [1., 1., 1., 0.],
              [1., 1., 1., 1.]])

...决定是否应显示散点图矩阵的每个子图的数据。如果 value = 0，则子图中的特定轨迹...

Histogram({
    'marker': {'color': 'rgb(31, 119, 180)'},
    'showlegend': False,
    'x': [-1.4052399956918005, -1.8538677019305498, -0.0016298185761457061,
          -0.5268239747464603, -0.10652357762295094, 0.02921346151566477,
          -0.4581921321144815, 0.020997665069043978, 0.8734380864952216,
          -1.3008441288553083],
    'xaxis': 'x',
    'yaxis': 'y'
})

... 简单地替换为:

go.Scatter()

如果子图中的所有轨迹都发生这种情况，那么该图将被空白，就像您在此处看到的那样:

plotly :

完整代码:

import plotly.graph_objects as go
import plotly.figure_factory as ff
import numpy as np
import pandas as pd
import plotly.express as px

cols = ['Column A', 'Column B', 'Column C', 'Column D']
# cols = ['Column A', 'Column B', 'Column C']
# cols = ['Column A', 'Column B', 'Column C', 'Column D', 'Column E', 'Column F']
df = pd.DataFrame(np.random.randn(20, len(cols)),
                columns=cols)

df['Fruit'] = pd.Series(['apple', 'apple', 'grape', 'apple', 'apple',
                                'grape', 'pear', 'pear', 'apple', 'pear',
                                'apple', 'apple', 'grape', 'apple', 'apple',
                                'grape', 'pear', 'pear', 'apple', 'pear'])

fig = ff.create_scatterplotmatrix(df, diag='histogram', index='Fruit',
                                  height=800, width=800)


# get number of rows and columns
ix = df['Fruit'].unique()
xs = []
ys = []
fig.for_each_xaxis(lambda x: xs.append(x.title.text) if x.title.text != None  else ())
fig.for_each_yaxis(lambda y: ys.append(y.title.text) if y.title.text != None  else ())

# build matrix to determine visiblity of trqces
m = (len(xs), len(ys))
show = np.tril(np.ones(m))
newshows = []
newdata = []
for i in range(0, len(show[0])):
    lst = list(np.repeat(show[i], len(ix)))
    newshows.extend(lst)

# replace existing data with empty data in the upper triangle
for t, s in enumerate(newshows):
    if  newshows[t] == 0:
        newdata.append(go.Scatter())
    else:
        newdata.append(fig.data[t])
        
fig.data = []
for i, new in enumerate(newdata):
    fig.add_trace(newdata[i])

fig.show()