Attempting to construct a plt.subplot(nrows=2,ncols=3) stacked, grouped bar chart on a DataFrame with a MultiIndex

area_fips    object
year          int32
qtr           int32

where col sector (for color) is a the first 2-digits of industry_code (for bar label).

Cols

qtrly_estabs           int32
total_qtrly_wages      int64
avg_mthly_emp        float64

are used to compute

estabs_rank          float64
wages_rank           float64
emp_rank             float64

where only the top 3 industry_code of each domain are graphed; their values are

pct_total_wages      float64
pct_total_emp        float64
pct_total_estabs     float64

year is a group of 4 qtr, which is a stack of top 3 industry_code, colored by sector.

subplot cols are qtrly_estabs, total_qtrly_wages, avg_mthly_emp

subplot rows are data point and pct

area_fips index level can be dropped, as a list of subplots will be made for the corresponding area_fips

sharex: True for all subplots, sharey: True for plots in second row.

qtr does not need to be visually delineated as in Create a grouped bar chart with Matplotlib and pandas

Also referenced Grouped bar chart with labels, Stacked and Grouped Bar Plot, and How to have clusters of stacked barsbut still can’t seem to structure my input properly.