How to modify figure with subplots using pandas?-CodePudding

Whenever I'm making subplots, I'm used to making each plot individually and combining them all in a list.

Pandas is really cool because, with just one simple line of code, I can make a figure that contains histograms of all the features of a dataset:

hist_plot = iris.hist(alpha=0.4, figsize=(10,6))

However, I'm stumped on how to make changes to this figure. For example:

Making each plot a different color: color = ['blue', 'red', 'green', 'purple']
Giving each plot a different title: title = ['SLength', 'SWidth', 'PLength', 'PWidth']
Adding "Frequency" label to each y axis

Is there a way to do this without having to make each plot individually? I tried this as an example, but it didn't work:

hist_plot = iris.hist(color = ['blue', 'red', 'green', 'purple'], alpha=0.4, figsize=(10,6))

CodePudding user response：

df.hist() returns a two-dimensional array of axes objects, which correspond to the subplots in their grid arrangement. To iterate over them, their flat attribute is convenient, as it contains the same objects in a flat array. You can use zip() to iterate over the subplots and your color and title lists in parallel:

from sklearn.datasets import load_iris

iris = load_iris(as_frame=True)['data']

colors = ['blue', 'red', 'green', 'purple']
titles = ['SLength', 'SWidth', 'PLength', 'PWidth']

axs = iris.hist(alpha=0.4, figsize=(10, 6))

for ax, color, title in zip(axs.flat, colors, titles):
    for patch in ax.patches:
        patch.set_color(color)
    ax.set_title(title)
    ax.set_ylabel('Frequency')