Dear listmembers,
I experienced a problem with matplotlib yesterday that caused me a lot of
headache and I don't know whether this is a bug or if I forgot something.
What I want is to plot different datasets in different plots in different
files. My code looks like that:
···
#######################
from pylab import fft, ifft, var, average, real, show
import matplotlib.pyplot as plt
def process_file(filename, fig_nr):
plt.figure(fig_nr)
plt.clf()
plt.subplot(211)
plt.plot(s, alt, '-')
plt.subplot(212)
plt.plot(s,map(lambda x: x*3.6,speed))
plt.savefig(filename+".png")
process_file("foo1.tcx", 1)
process_file("foo2.tcx", 2)
process_file("foo3.tcx",3)
##########################
The problem here is that in the plots for file foo2 the data from foo1 is
included, in the plots for foo3 the data of foo1 and foo2.
I thought that I paid respect to this behaviour with plt.figure(), on
stackoverflow I read about the plt.clf() - thing but nothing happened. I tried
to make fig_nr a hash over the filename or leave it blank (like suggested in
the documentation) but this had no effect and I ended up with the additional
parameter (fig_nr in the example above) which also did not work.
Is this maybe a bug or did I forget an important thing?
I'm running Debian sid with python-matplotlib 0.99.1.2-2 (I hope the numbering
from Debian and matplotlib is similar).
Best regards,
Philipp Bender