Reputation: 2007
I would like to create one pdf file with 12 plots, in two options:
Using plt.savefig("months.pdf")
saves only last plot.
MWE:
import pandas as pd
index=pd.date_range('2011-1-1 00:00:00', '2011-12-31 23:50:00', freq='1h')
df=pd.DataFrame(np.random.randn(len(index),3).cumsum(axis=0),columns=['A','B','C'],index=index)
df2 = df.groupby(lambda x: x.month)
for key, group in df2:
group.plot()
I also tried:
fig, axes = plt.subplots(nrows=2, ncols=2, figsize=(15, 10))
after the group.plot
but this produced four blank plots...
I have found an example of PdfPages but I don't know how to implement this.
Upvotes: 2
Views: 2859
Reputation: 40973
To save a plot in each page use:
from matplotlib.backends.backend_pdf import PdfPages
# create df2
with PdfPages('foo.pdf') as pdf:
for key, group in df2:
fig = group.plot().get_figure()
pdf.savefig(fig)
In order to put 4 plots in a page you need to first build a fig with 4 plots and then save it:
import matplotlib.pyplot as plt
from itertools import islice, chain
def chunks(n, iterable):
it = iter(iterable)
while True:
chunk = tuple(islice(it, n))
if not chunk:
return
yield chunk
with PdfPages('foo.pdf') as pdf:
for chunk in chunks(4, df2):
fig, axes = plt.subplots(nrows=2, ncols=2, figsize=(12, 4))
axes = chain.from_iterable(axes) # flatten 2d list of axes
for (key, group), ax in zip(chunk, axes):
group.plot(ax=ax)
pdf.savefig(fig)
Upvotes: 1