pandas dataframe binned data plotting

Question

I have a pandas dataframe (df) with a column df.age ranging from 0 - 100 years and a column df.haircolor (bright, dark, purple, grey).

Now I want to bin the age in decades:

bins = np.linspace(df.age.min(), df.age.max(), 10)
decades = df.groupby(np.digitize(df.age, bins))

Now I am trying to find a good way to plot this. I'd like a barplot where each haircolor has a bar. Naively I tried that.

df['haircolor'].plot(kind='bar', by=decades)

It is not giving me the result I hoped for. Anyone an idea? thanks.

Alexander · Answer 1 · 2015-05-18T15:37:49.227

2

Try this:

df['decade'] = df.age // 10 * 10
df.groupby(['decade', 'haircolor']).haircolor.count().plot(kind='bar')

edited May 18 '15 at 15:37

answered May 17 '15 at 22:52

Alexander

That gives me a KeyError: 'decades' – Annamarie May 18 '15 at 07:32
Sorry, I thought you already had included the decade in your DataFrame. See above for calculation. – Alexander May 18 '15 at 15:38
Thanks Alexander, that works - do you know a way to colorcode the different haircolors too? – Annamarie May 18 '15 at 16:05
http://stackoverflow.com/questions/11927715/how-to-give-a-pandas-matplotlib-bar-graph-custom-colors – Alexander May 18 '15 at 16:12

1 Answers1