如何为 Plotly 条形图设置 static x 轴值和图例集？

Question

The Plotly Bar graphs below show a count of employees in three different seniority classes for a given dept (engineering, accounting, etc..)

If one of the many values for the x-axis is not in the current data frame used to construct the graph, no bar/x-axis/color is shown. Below accounting has employees in all three so they show but accounting doesn't possess all there.

Is there something I can do to make sure this graph consistently shows all three values and that a missing x-axis value will assume a default y-axis value of 0.

I want the engineering graph to show an bar for sr as well but default the value to 0.

Ploty Graph showing Count by Seniority (Junior, Mid, Senior) for Engineering Dept

Ploty Graph showing Count by Seniority (Junior, Mid, Senior) for Accounting Dept

Here is a snippet of the code use to generate the graph above as well as snapshot of the master Dataframe.

fig = px.bar(g[g['dept']=='accounting']
         , x='level', y='id', color='level', labels = {
             'level' : 'Job Seniority',
             'id' : 'Employee Count'
         }, title='Count by Seniority (Junior, Mid, Senior) for Accounting Dept')

In code snippet above, g points to a Dataframe that looks like this:

	dept	level	id
0	accounting	jr	1
1	accounting	mid	2
2	accounting	sr	1
3	business development	jr	2
4	business development	mid	1
5	business development	sr	1
6	engineering	jr	2
7	engineering	mid	1
8	human resources	jr	1
9	legal	jr	1
10	legal	mid	1
11	legal	sr	1
12	marketing	jr	2
13	marketing	mid	1
14	research and development	jr	1
15	sales	sr	1
16	services	jr	2
17	services	mid	1
18	support	jr	3
19	support	mid	2
20	training	mid	2

Answer 1

• prepare the data for plotting by organising index and columns
• used graph_objects so make_subplots() can be used so it's a simple looping process to generate a graph per department (not plotly is 1-indexed, python 0-indexed)

import numpy as np
import pandas as pd
import plotly.graph_objects as go
from plotly.subplots import make_subplots


df = pd.DataFrame({"dept": ["accounting", "accounting", "accounting", "business development", "business development", "business development", "engineering", "engineering", "human resources", "legal", "legal", "legal", "marketing", "marketing", "research and development", "sales", "services", "services", "support", "support", "training"], 
 "level": ["jr", "mid", "sr", "jr", "mid", "sr", "jr", "mid", "jr", "jr", "mid", "sr", "jr", "mid", "jr", "sr", "jr", "mid", "jr", "mid", "mid"], 
 "id": [1, 2, 1, 2, 1, 1, 2, 1, 1, 1, 1, 1, 2, 1, 1, 1, 2, 1, 3, 2, 2]})

dfp = df.groupby(["dept","level"]).sum().unstack("level").fillna(0).droplevel(0,1).T

fig = make_subplots(rows=(len(dfp.columns)+1)//3, cols=3, subplot_titles=dfp.columns)
for i, c in enumerate(dfp.columns):
    fig.add_trace(go.Bar(x=dfp.index, y=dfp[c], showlegend=False, 
                         marker_color=np.select([dfp.index=="jr",dfp.index=="mid"],["red","blue"],"green")), 
                  row=(i//3)+1, col=(i%3)+1)
    
fig.update_layout({"margin":{"b":0,"t":20,"l":0,"r":0}})