I'm trying to use pandas to plot to as specific figure, but it seems to want to make it's own figures and not use / resets pyplot's current figure.
How can I make pandas plot to the current (or better yet, and explicitly given) figure?
from matplotlib import pyplot
import pandas
df = pandas.DataFrame({'a': list(range(1000))})
fig1 = pyplot.figure(figsize=(5, 10))
assert pyplot.gcf() is fig1 # succeeds
df.plot() # does not draw to fig1
assert pyplot.gcf() is fig1 # fails
The solution is quite easy. You can specify to which axes object the plot needs to refer. This can be done by getting the current handle from ax = pyplot.gca(), subsequently plotting to this handle. Of course, you can always plot to another handle as well using similar syntax.
from matplotlib import pyplot
import pandas
df = pandas.DataFrame({'a':list(range(1000))})
fig1 = pyplot.figure(figsize=(5, 10))
ax = pyplot.gca()
assert pyplot.gcf() is fig1 # succeeds
df.plot(ax=ax) # draws to fig1 now
assert pyplot.gcf() is fig1 # Succeeds now, too
df = pandas.DataFrame(range(1000)) results in: Traceback (most recent call last): ... .... ... ValueError: DataFrame constructor not properly called!
range to DataFrame. This is more an issue of using python 3 versus python 2. In python 3 you'd need to convert to list first, pd.DataFrame(list(range(1000))).Not exactly your question, but may address your need via different mechanism:
If you just want to set the figure size, you can pass pyplot kwargs to df.plot:
df.plot(figsize=(5, 10))
pyplot.show()
