How to plot a stacked bar chart from a pandas frame which is grouped and counted

Question

My goal is to produce a simple stacked bar plot with the VALUE as the X axis, each type is the bar stacks, and the number(count) is the y value of each bar stack.

I have the result of a data frame grouping which looks like:

VALUE	TYPE
0.0	0	16
	1	26
	2	15
	3	18
1.0	1	15
	2	2
	3	10
2.0	0	10
	1	175
	2	99
	3	11
	15	100
3.0	0	92
	1	20
	2	17
	3	26
	15	1

Name: VALUE, dtype: int64

I created this from a large data frame by: dataOutput = df.groupby(['VALUE','TYPE']).agg('VALUE').count()

I want to iterate through the dataframe and get access to the 'counts', and then push the uneven result into the stacked plots.

I am stuck at the looping through this data and getting access to the count. for i, (val, typ) in enumerate(df.index): print(val) print(typ)

How to access count?

for val, typ in df.groupby(level=0): print(val) print(typ) break

The data is there, but not sure how to access the count column.

VALUE	TYPE
0.0	0	16
	1	26
	2	15
	3	18

Is there some simpler way? Or can someone explain how I could condition this data.

Answer 1

IIUC, you have a multiindex series, s, that looks like this:

VALUE  TYPE
0.0    0        16
       1        26
       2        15
       3        18
1.0    1        15
       2         2
       3        10
2.0    0        10
       1       175
       2        99
       3        11
       15      100
3.0    0        92
       1        20
       2        17
       3        26
       15        1
Name: count, dtype: int64

For convenience, I renamed the series name to 'count'.

s.unstack().plot.bar(stacked=True)

Output: