I have a sample dataframe like the following:
import pandas as pd
df = pd.DataFrame(np.random.randint(
0, 10, size=(1000, 11)), columns=list('ABCDEFGHIJK'))
The desired but unpolished output looks like this:
The data of each column in the dataframe is plotted as a subplot with five rows of bars.
I prefer to use matplotlib because I can relatively easily make the graphs looking good. But its performance seems pretty slow.
You can use the bottom parameter of bar to offset the individual rows.
The following not optimized example demonstrates this approach:
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
df = pd.DataFrame(np.random.randint(0, 10, size=(1000, 11)), columns=list('ABCDEFGHIJK'))
fig = plt.figure()
for i,c in enumerate(df.columns):
ax = fig.add_subplot(3, 4, i+1)
for x,h,b in zip((df.index.to_numpy() % 200).reshape(-1, 200), df[c].to_numpy().reshape(-1, 200), (df.index.to_numpy() // 200 * 10).reshape(-1, 200)):
ax.set_title(c)
ax.bar(x, h, bottom=b, color='k' )
