There are a few things that I would like to express in a bar chart and have no clue about doing it using basic graphing techniques in matplotlib. I have a dataframe which is shown below and would like to obtain a bar chart as described below. The x-axis is based on the the Type column of the dataframe and within a single bar, the different colors are based on the Name column and the size of the bar is defined by the Count number. The color of the different names need not to be the same across different types,
as long as the colors within a single bar is different.
Add a comment
|
2 Answers
You can use pivot and then plot
df.pivot('Type', 'Name', 'Count').plot(kind = 'bar', stacked = True, color = ['b','g','orange','m', 'r'])
Edit: To sort the values
df.pivot('Type', 'Name', 'Count').sort_values(by = 'A', ascending = False, axis = 1)\
.plot(kind = 'bar', stacked = True, color = ['g','r','b','orange', 'm'])
3 Comments
Stanley Gan
Is there a way to sort the bar chart in descending order of the counts?
Stanley Gan
I think I know how. I can create an extra column called 'Sum' which is the sum of the other columns and sort based on that before plotting it. Thanks a lot for your help! This works great :)
Vaishali
See the edit for sort and thank you for accepting :)
I you want to change the size of plot the use arg figsize=(15, 5)
df.pivot('Type', 'Name', 'Count').sort_values(by = 'A', ascending = False, axis = 1)\
.plot(kind = 'bar', stacked = True, color = ['g','r','b','orange', 'm'], figsize=(15,5))
Comments
