Runtime Error - Need Advice

Lee Boger wrote:

Thanks for the quick feedback. I changed the last line from figure.figure.show() to pyplot.show() and it worked without crashing the interpreter. However, if I close the figure then re-run the script, a new figure pops up but it doesn't have any data plotted. Interpreter is now locked up, the run-time error occurred and no figure stored as a file log_plot. I un-installed numpy-1.4.0 and installed numpy-1.3.0, but the results were as I just described.

Lee

At this point a Windows user may need to step in. I know nothing about PythonWin. My suspicion is that this is a problem of dueling event loops, and one solution would be to use ipython instead of pythonwin.

(Note that you typically need to use "reply to all" on this list, otherwise the reply does not get copied to the list.)

Eric

···

*Eric Firing <efiring@...202...>*

02/09/2010 01:35 PM

To
  Lee Boger <Boger_Lee@...2968...>
cc
  matplotlib-users@lists.sourceforge.net
Subject
  Re: [Matplotlib-users] Runtime Error - Need Advice

Caterpillar: Confidential Green Retain Until: 03/11/2010

Lee Boger wrote:
>
> Windows XP Professional with Python 2.5 installed (pywin32 build 210) -
> came with dSPACE software package
>
> Downloaded and installed matplotlib-0.99.1.win32-py2.5.exe from
> sourceforge.net
>
> Downloaded and installed numpy-1.4.0-win32-superpack-python2.5.exe from
> sourceforge.net

I suspect this is the problem: numpy-1.4.0 is considered broken and has
been withdrawn. Specifically, it introduces binary incompatibility with
programs compiled against prior versions of numpy, including matplotlib.
If you go back to the numpy sourceforge site now, I think you will see
an earlier version that you can install in place of 1.4.0 (after
removing the latter).

Eric

>
> Executing the following simple "log plot" script within PythonWin:
>
> *from* matplotlib *import* pylab
>
> # Create some artificial data.
> test_frequency = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16,
> 17, 18, 19, 20]
> test_results = [-0.2, -0.7, -1.0, -1.5, -2.0, -2.5, -3.0, -3.5, -4, -5,
> -6, -7.1, -8, -9, -10, -11, -12, -15, -20, -25]
> spec_frequency = [6, 8]
> spec_results = [-3.0, -3.0]
>
> # Plot
> figure = pylab.subplot(111)
> figure.semilogx()
> figure.scatter(test_frequency, test_results, s=20, c='b', marker='s',
> edgecolors='none')
> figure.scatter(spec_frequency, spec_results, s=40, c='g', marker='s',
> edgecolors='none')
> figure.grid(True)
> figure.set_xlabel(r"Frequency (Hz)", fontsize = 12)
> figure.set_ylabel(r"Actuator Response (db)", fontsize = 12)
>
> figure.figure.savefig('log_plot')
> figure.figure.show()
>
> Plots a figure on the screen that looks correct, then the following
> error (when I click OK, PythonWin closes)
>
> Any advice would be appreciated. Maybe it's an installation or setup
> issue, but I'm pretty knew to Python programming and don't know how to
> debug this.
>
> Lee Boger

Thanks for your help. I’ll look into
ipython.

Lee

Eric Firing <efiring@…202…>
02/09/2010 02:31 PM

To

Lee Boger <Boger_Lee@…2968…>,
matplotlib-users matplotlib-users@lists.sourceforge.net
cc

Subject

Re: [Matplotlib-users] Runtime Error

  • Need Advice

Caterpillar: Confidential Green
Retain Until: 03/11/2010

`Lee Boger wrote:

Thanks for the quick feedback. I changed the last line from

figure.figure.show() to pyplot.show() and it worked without crashing
the

interpreter. However, if I close the figure then re-run the script,
a

new figure pops up but it doesn’t have any data plotted. Interpreter
is

now locked up, the run-time error occurred and no figure stored as
a

file log_plot. I un-installed numpy-1.4.0 and installed numpy-1.3.0,
but

the results were as I just described.

Lee

At this point a Windows user may need to step in. I know nothing
about

PythonWin. My suspicion is that this is a problem of dueling event

loops, and one solution would be to use ipython instead of pythonwin.

(Note that you typically need to use “reply to all” on this list,

otherwise the reply does not get copied to the list.)

Eric

Eric Firing <efiring@…202…>

02/09/2010 01:35 PM

To

             Lee

Boger <Boger_Lee@…2968…>

cc

             matplotlib-users@lists.sourceforge.net

Subject

             Re:

[Matplotlib-users] Runtime Error - Need Advice

Caterpillar: Confidential Green Retain
Until: 03/11/2010

Lee Boger wrote:

Windows XP Professional with Python 2.5 installed (pywin32
build 210) -

came with dSPACE software package

Downloaded and installed matplotlib-0.99.1.win32-py2.5.exe
from

sourceforge.net

Downloaded and installed numpy-1.4.0-win32-superpack-python2.5.exe
from

sourceforge.net

I suspect this is the problem: numpy-1.4.0 is considered broken and
has

been withdrawn. Specifically, it introduces binary incompatibility
with

programs compiled against prior versions of numpy, including matplotlib.

If you go back to the numpy sourceforge site now, I think you will
see

an earlier version that you can install in place of 1.4.0 (after

removing the latter).

Eric

Executing the following simple “log plot” script
within PythonWin:

from matplotlib import pylab

Create some artificial data.

test_frequency = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13, 14, 15, 16,

17, 18, 19, 20]

test_results = [-0.2, -0.7, -1.0, -1.5, -2.0, -2.5, -3.0,
-3.5, -4, -5,

-6, -7.1, -8, -9, -10, -11, -12, -15, -20, -25]

spec_frequency = [6, 8]

spec_results = [-3.0, -3.0]

Plot

figure = pylab.subplot(111)

figure.semilogx()

figure.scatter(test_frequency, test_results, s=20, c=‘b’,
marker=‘s’,

edgecolors=‘none’)

figure.scatter(spec_frequency, spec_results, s=40, c=‘g’,
marker=‘s’,

edgecolors=‘none’)

figure.grid(True)

figure.set_xlabel(r"Frequency (Hz)", fontsize
= 12)

figure.set_ylabel(r"Actuator Response (db)",
fontsize = 12)

figure.figure.savefig(‘log_plot’)

figure.figure.show()

Plots a figure on the screen that looks correct, then the
following

error (when I click OK, PythonWin closes)

Any advice would be appreciated. Maybe it’s an installation
or setup

issue, but I’m pretty knew to Python programming and don’t
know how to

debug this.

Lee Boger

`