CODEWITHSUNDEEP
pythonpandasmatplotlib
Vitek Rous
Vitek Rous

Reputation: 45

Pandas boxplot, different y axes in subplots

First, to say. I'm aware of similar questions namely pandas-boxplot-groupby-different-ylim-in-each-subplot and different-ylim-for-shared-axes-in-pandas-boxplot but I am still not able to get the desired result with my dataset. I've got a dataset of wastewater samples where columns are different pollutants and rows with "codes" for three different groups:

Dataset looks like this (but bigger)

data = pd.DataFrame({'Group': pd.Series(['IN', 'IN', 'IN', 'OUT', 'OUT', 'OUT']),
                      'COD' : pd.Series([800, 900, 100, 400, 450, 500]),
                      'BOD' : pd.Series([40, 50, 60, 20, 25, 30])})

and my code for boxplot (in jupyter with matplotlib inline)

fig, ax = plt.subplots(figsize=(10,  5))
data.boxplot(['COD','BOD'], 'Group', ax)

and the output is like this

boxplot output

It's obvious that the y-axis can't be the same for all the data, but I was not able to recreate the output in such way to change the axis for individual subplots even with help of the links above. I would really a appreciate any help on this.

Upvotes: 4

Views: 7600

Answers (2)

Diziet Asahi
Diziet Asahi

Reputation: 40697

You need to create your axes using sharey=False before calling DataFrame.boxplot() to force the non-sharing of y-axes limits. But you have to be careful to provide the correct number of axes, otherwise pandas is forced to clear the figure and create it's own axes (which will then share their y-axes).

fig, axs = plt.subplots(1,len(data.Group.unique()),figsize=(10,  5), sharey=False)
data.boxplot(['COD','BOD'], 'Group', axs)

enter image description here

Upvotes: 8

Scott Boston
Scott Boston

Reputation: 153460

One way you could do this create different axes for the subplots and plotting the dataframe columns separately:

fig, ax = plt.subplots(1, 2, figsize=(10, 5))
data.boxplot('COD','Group', ax=ax[0])
data.boxplot('BOD','Group', ax=ax[1])

Output:

enter image description here

Upvotes: 7

Related Questions