plotly 的堆疊和分組條形圖

Question

我有這個數據集

import pandas as pd
import plotly.express as px

elements = pd.DataFrame(data={"Area": ["A", "A", "A", "B", "B", "C", "C", "C"], "Branch": ["a1", "f55", "j23", "j99", "ci2", "p21", "o2", "q35"], "Good": [68, 3, 31, 59, 99, 86, 47, 47], "Neutral": [48, 66, 84, 4, 83, 76, 6, 89],"Bad": [72, 66, 50, 83, 29, 54, 84, 55]})

  Area Branch  Good  Neutral  Bad
0    A     a1    68       48   72
1    A    f55     3       66   66
2    A    j23    31       84   50
3    B    j99    59        4   83
4    B    ci2    99       83   29
5    C    p21    86       76   54
6    C     o2    47        6   84
7    C    q35    47       89   55

我正在嘗試 plot 它並得到看起來像這樣的東西

堆疊並與標簽分組，所以我嘗試了這個

fig_elements = px.bar(elements, x='Branch', y=["Good", "Neutral", "Bad"], orientation="v", template="plotly_dark", facet_col='Area')
fig_elements.update_layout(plot_bgcolor="rgba(0,0,0,0)", xaxis=(dict(showgrid=False)), yaxis=(dict(showgrid=False)), barmode="stack")

無法在條形圖上添加標簽和區域，並且區域具有所有分支，而不僅僅是屬於它的分支，我該如何解決這個問題並添加標簽？

Answer 1

import plotly.graph_objects as go
df = elements.melt(id_vars=['Area', 'Branch'], value_vars=['Good', 'Neutral', 'Bad'], var_name='Rating')
df
###
   Area Branch   Rating  value
0     A     a1     Good     68
1     A    f55     Good      3
2     A    j23     Good     31
3     B    j99     Good     59
4     B    ci2     Good     99
5     C    p21     Good     86
6     C     o2     Good     47
7     C    q35     Good     47
8     A     a1  Neutral     48
9     A    f55  Neutral     66
10    A    j23  Neutral     84
11    B    j99  Neutral      4
12    B    ci2  Neutral     83

Plot

rating_color = ['#17BEBB', '#FFD97D', '#EE6055']
fig = go.Figure()
fig.update_layout(barmode='stack', width=900, height=800, uniformtext_minsize=8, uniformtext_mode='hide')
for r, c in zip(df['Rating'].unique(), rating_color):
    df_plot = df[df['Rating'] == r]
    fig.add_trace(
        go.Bar(
            x=[df_plot['Area'], df_plot['Branch']],
            y=df_plot['value'],
            marker_color=c,
            name=r,
            text=df_plot['value'],
            width=0.45,
            textposition='auto')
    )

fig.show()