seaborn multiple variables group bar plot

Question

I have pandas dataframe, one index(datetime) and three variables(int)

date        A   B       C
2017-09-05  25  261     31
2017-09-06  261 1519    151
2017-09-07  188 1545    144
2017-09-08  200 2110    232
2017-09-09  292 2391    325

I can create grouped bar plot with basic pandas plot.

df.plot(kind='bar', legend=False)

However, I want to display in Seaborn or other libraries to improve my skills.
I found very close answer(Pandas: how to draw a bar plot with two categories and four series each?).
In its suggested answer, it has code that

ax=sns.barplot(x='', y='', hue='', data=data)

If I apply this code to mine, I do not know what my 'y` would be.

ax=sns.barplot(x='date', y=??, hue=??, data=data)

How can I plot multiple variables with Seaborn or other libraries?

Solution

I think need melt if want use barplot:

data = df.melt('date', var_name='a', value_name='b')
print (data)
          date  a     b
0   2017-09-05  A    25
1   2017-09-06  A   261
2   2017-09-07  A   188
3   2017-09-08  A   200
4   2017-09-09  A   292
5   2017-09-05  B   261
6   2017-09-06  B  1519
7   2017-09-07  B  1545
8   2017-09-08  B  2110
9   2017-09-09  B  2391
10  2017-09-05  C    31
11  2017-09-06  C   151
12  2017-09-07  C   144
13  2017-09-08  C   232
14  2017-09-09  C   325

ax=sns.barplot(x='date', y='b', hue='a', data=data)
ax.set_xticklabels(ax.get_xticklabels(), rotation=90)

Pandas solution with DataFrame.plot.bar and set_index:

df.set_index('date').plot.bar()

Comments

[jayko03]

I tried, but I got KeyError:datetime.date(2017, 9,5). And I don't think melt` is what I want. I just want to express graph(photo in my question) to more pretty and visualized with Seaborn or other libraries.

[jezrael]

What version of seaborn so you use?

[jayko03]

I am using 0.8.1 version

[jezrael]

@jaykodeveloper - I use old version, give me a some time for upgrade and test.

[jezrael]

I upgrade and it working nice, maybe problem is in pandas version, I use 0.22.0

[jayko03]

I use pandas 0.22.0. Okay. My issue is that when I pass date as first parameter, it throws KeyError. My df.index is Index([2017-09-05, 2017-09-06, 2017-09-07, 2017-09-08, 2017-09-09, 2017-09-10, 2017-09-11, 2017-09-12, 2017-09-13, 2017-09-14, 2017-09-15, 2017-09-16, 2017-09-17, 2017-09-18, 2017-09-19, 2017-09-20, 2017-09-21, 2017-09-25, 2017-09-26, 2017-09-27, 2017-10-04], dtype='object', name='date') Is it how you made your dataframe?

[jayko03]

Let us continue this discussion in chat.

[jezrael]

You are really close data = df.reset_index().melt('date', var_name='a', value_name='b')

[Long Le]

This answer works nicely, if you plan to plot a dataframe with barplot, this is the answer