用辅助y轴绘制大熊猫列并将图表保存到单个pdf

Question

I am trying to create multiple time-series plots using data in a Pandas DataFrame columns. I want to save all these charts to a single pdf file. The challenge is that these charts all have a secondary y-axis which has a different scale (see Col Z below). I have seen examples of saving multiple charts to pdf, and examples of plotting a singe chart with a secondary y-axis, but don't know how to combine the two. My actual DataFrame has 50 columns, so doing it manually would be tedious. Thanks in advance.

Here is what I tried:

df1 = pd.DataFrame(np.random.rand(10, 4))
df1.columns = ['A', 'B', 'C', 'Z']
date_range = pd.date_range('1/31/2011', periods=10, freq='M')
df1['Z'] = 100*df1['Z']
df1.plot(kind='line', subplots=True, secondary_y=['Z'])
#

import matplotlib.backends.backend_pdf
pdf = matplotlib.backends.backend_pdf.PdfPages("output.pdf")
plt.figure()
for cols in df1.columns:
    df1.plot(y=cols, secondary_y=['Z'])
    pdf.savefig(fig)
pdf.close()

Answer 1

As I understand, I think this is what you are trying. This generate 4 graphs with 2 series in each one. Series A , B , C are plotted together with serie Z :

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.backends.backend_pdf

df1 = pd.DataFrame(np.random.rand(10, 4))
df1.columns = ['A', 'B', 'C', 'Z']
date_range = pd.date_range('1/31/2011', periods=10, freq='M')
df1['Z'] = 100*df1['Z']
df1.plot(kind='line', subplots=True, secondary_y=['Z'])

with matplotlib.backends.backend_pdf.PdfPages("output.pdf") as pdf:
    for cols in df1.columns:
        ## plot 1
        fig, ax1 = plt.subplots()
        df1.plot(y=cols, ax=ax1, color='C1')
        ax1.legend(loc=2)
        ## plot 2
        ax2 = ax1.twinx()
        df1.plot(y='Z', ax=ax2)
        pdf.savefig()