CODEWITHSUNDEEP
pythonpandasplot
Mike
Mike

Reputation: 643

Why is Pandas plotting index values (x_ticklabels) incorrectly?

I've got this data frame

                        Poverty      White      Black  Native_American
CHSI_State_Name                                                          
Alabama               17.085075  69.795522  28.510448         0.570149   
Alaska                12.144444  54.485185   1.174074        36.159259   
Arizona               16.840000  81.086667   1.886667        14.813333   
...

And I plot my figure like this:

fig, ax = plt.subplots()
poverty.plot(ax = ax, figsize = (16, 8))
ax.set_xticklabels(poverty.index, rotation = 90)

fig.savefig('output7.png', xbbox_inches = 'tight')

I would expect the file to have state names under each value (all states), but instead it's only got a few with large gaps.Output graph

What am I doing wrong? Thanks!

Upvotes: 3

Views: 156

Answers (3)

Niki van Stein
Niki van Stein

Reputation: 10724

The problem in your code is that 

ax.set_xticklabels(poverty.index, rotation = 90)

sets the xticklabels on the range of the data. If your data ranges from 10 to 16, it will put names on 10,11,12,13,14 and 15. Instead of putting labels on each integer xtick you want to specify the position / order of the labels as well.

By using 

ax.set_xticks(range(len(poverty.index)))

You specify the range as big as the poverty.index which makes sure that each label is positioned inside your plot.

Upvotes: 1

Mike
Mike

Reputation: 643

This seems to do it:

ax.set_xticklabels(poverty.index, rotation = 90)
ax.set_xticks(range(len(poverty.index)))

Upvotes: 1

Fabio Lamanna
Fabio Lamanna

Reputation: 21542

Try substitute ax.set_xticklabels(poverty.index, rotation = 90) with:

locs, labels = plt.xticks()
plt.setp(labels, rotation=90)

For me it works with the expected behaviour. But actually I don't know how your method does not work. This is a proposal workaround. Hope that helps.

Upvotes: 0

Related Questions