There is definitely something wrong with matplotlib/numpy. Consider the
following
>from numpy import *
>mydata=memmap('map.dat',dtype=float64,mode='w+',shape=56566500)
> del mydata
I can now remove the file map.dat with (from the command line) $rm map.dat
However
If I plot mydata before the line
> del mydata
I can't get rid of the file until I exit python!!
Does matplotlib keep a reference to the data? How can I remove this
reference?
Mathew
Here is sample code demonstrating the problem I'm having with
pyhd5/numpy/matplotlib
import h5py
import numpy
import re
import sys
import os
import gtk
import pdb
import gc
from matplotlib.figure import Figure
from matplotlib.backends.backend_gtkagg import FigureCanvasGTKAgg as
FigureCanvas
def doplot(widget,box1):
id=h5py.h5f.open("MLS-Aura_L1BRADD_v02-23-c01_2009d001.h5")
data_id = h5py.h5d.open(id,"R1A:118.B22D:PT.S0.DACS-4")
dataspaceID = h5py.h5d.DatasetID.get_space(data_id)
noDims = h5py.h5s.SpaceID.get_simple_extent_ndims(dataspaceID)
dims = h5py.h5s.SpaceID.get_simple_extent_dims(dataspaceID)
start=tuple([0]*noDims)
h5py.h5s.SpaceID.select_hyperslab(dataspaceID,start,dims,None,None,
h5py.h5s.SELECT_SET)
memory_space_id = h5py.h5s.create_simple(dims)
data=numpy.memmap('tmpdat',dtype=numpy.float64,mode='w+',shape=dims)
h5py.h5d.DatasetID.read(data_id,memory_space_id,dataspaceID,data)
del data
data=numpy.memmap('tmpdat',dtype=numpy.float64,mode='r',shape=dims)
fig = Figure(figsize=(5,5), dpi=100)
ax = fig.add_subplot(111)
plot_data=data[0,0:,0]
mif=numpy.arange(plot_data.shape[0])
#if the next line is commented out, all is good
ax.plot(plot_data,mif)
canvas = FigureCanvas(fig)
box1.pack_start(canvas, True, True, 0)
canvas.show()
id.close()
del data
del fig
del ax
del plot_data
del canvas
gc.collect()
#the following line fails
os.remove('tmpdat')
def delete_event(widget, event, data=None):
return False
window = gtk.Window(gtk.WINDOW_TOPLEVEL)
window.connect("destroy", lambda x: gtk.main_quit())
box1 = gtk.HBox(False, 0)
window.add(box1)
button = gtk.Button("Hello World")
box1.pack_start(button, True, True, 0)
#window.add(box1)
button.show()
button.connect("clicked", doplot, box1)
box1.show()
window.set_default_size(500,400)
window.show()
gtk.main()
Yes, for efficiency reasons matplotlib will use a reference to original
data whenever possible, rather than copying it. Also, when using the
pylab/pyplot API, matplotlib figures also stay around until they are
explicitly deleted. You may need an explicit call to "clf('all')" to
remove the figures, and thus references to the memmap'd file.
However, my advice would be to just not use memmap. I can't imagine
matplotlib performs very well if the data can't all fit in memory, so
you might as well just load it all into memory anyway.
Mike
Mathew Yeates wrote:
> There is definitely something wrong with matplotlib/numpy. Consider
> the following
> >from numpy import *
> >mydata=memmap('map.dat',dtype=float64,mode='w+',shape=56566500)
> > del mydata
>
> I can now remove the file map.dat with (from the command line) $rm map.dat
>
> However
> If I plot mydata before the line
> > del mydata
>
>
> I can't get rid of the file until I exit python!!
> Does matplotlib keep a reference to the data? How can I remove this
> reference?
>
> Mathew
> ------------------------------------------------------------------------
>
> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
> trial. Simplify your report design, integration and deployment - and focus on
> what you do best, core application coding. Discover what's new with
> Crystal Reports now. http://p.sf.net/sfu/bobj-july
> ------------------------------------------------------------------------
>
> _______________________________________________
> Matplotlib-devel mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>
--
Michael Droettboom
Science Software Branch
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA