Another Gnuplot style question

Joseph_Smidt · April 29, 2009, 6:20am

Okay, I am another gnuplot user trying to migrate over to matplotlib.
I like what I see, but there are a couple things that are very easy to
do in Gnuplot that I can't figure out how to do with matplotlib.

I have a file with 3 columns of data called data.txt that looks like:

0.0000 1.0000 1.0
0.0634 1.0655 1.1353
0.1269 1.1353 1.28899916094
0.1903 1.2097 1.46345358199
0.2538 1.2889 1.6615188369
0.3173 1.3734 1.88639043926
...

I can plot this data, 2 versus 1 and 3 versus 1, very easily on the
same plot, with a legend, with log y values, and only for the xrange
between 2 and 3 with gnuplot:

set log y
set xrange[2:3]
plot 'data.txt' u 1:2 w l t 'apples', 'data.txt' u 1:3 w l t 'oranges'

Now, how do I do that same thing with matplotlob? Ie:

1. Both graphs overlayed on the same plot.
2. Semilogy. (log y values),
3. Only ploy for x in the range 2-3.
4. Legend for the two graphs on same plot.

I have spent time looking through the documentation but I can't find
anyway to do this is any straightforward way. plotfile() looks
promising, but I can't seem to make it do the above. Thanks in
advance.

Joseph Smidt

···

--
------------------------------------------------------------------------
Joseph Smidt <josephsmidt@...287...>

Physics and Astronomy
4129 Frederick Reines Hall
Irvine, CA 92697-4575
Office: 949-824-3269

_Sandro_Tosi2 · April 29, 2009, 7:15am

1. Both graphs overlayed on the same plot.
2. Semilogy. (log y values),
3. Only ploy for x in the range 2-3.
4. Legend for the two graphs on same plot.

a simple example:

In [2]: import numpy as np

In [3]: import matplotlib.pyplot as plt

In [16]: data = np.random.rand(10,3)

In [17]: data
Out[17]:
array([[ 0.00669083, 0.4421283 , 0.46697081],
       [ 0.18093819, 0.11669917, 0.70887601],
       [ 0.11659791, 0.96514955, 0.07389404],
       [ 0.95616662, 0.30350482, 0.10036185],
       [ 0.14197553, 0.10560376, 0.2964961 ],
       [ 0.74705585, 0.21806946, 0.37095176],
       [ 0.1551145 , 0.76093425, 0.878701 ],
       [ 0.44315466, 0.3625146 , 0.06750168],
       [ 0.96109656, 0.88401174, 0.59215722],
       [ 0.46190334, 0.39079641, 0.5958516 ]])

In [28]: plt.semilogy(data[:,0], data[:,1], label='first dataset')
Out[28]: [<matplotlib.lines.Line2D object at 0xa7698ac>]

In [29]: plt.semilogy(data[:,0], data[:,2], label='second dataset')
Out[29]: [<matplotlib.lines.Line2D object at 0xa8049ac>]

In [30]: plt.xlim([0.2, 0.8]) # limit x to 0.2..0.8
Out[30]: (0.20000000000000001, 0.80000000000000004)

In [31]: plt.legend()
Out[31]: <matplotlib.legend.Legend object at 0xa80d04c>

In [32]: plt.show()

I have spent time looking through the documentation but I can't find
anyway to do this is any straightforward way. plotfile() looks
promising, but I can't seem to make it do the above. Thanks in
advance.

to load data from file you can try

matplotlib.mlab.csv2rec

Cheers,

···

On Wed, Apr 29, 2009 at 08:20, Joseph Smidt <josephsmidt@...287...> wrote:
--
Sandro Tosi (aka morph, morpheus, matrixhasu)
My website: http://matrixhasu.altervista.org/
Me at Debian: http://wiki.debian.org/SandroTosi

Eric_Firing1 · April 29, 2009, 7:20am

Joseph Smidt wrote:

Okay, I am another gnuplot user trying to migrate over to matplotlib.
I like what I see, but there are a couple things that are very easy to
do in Gnuplot that I can't figure out how to do with matplotlib.

I have a file with 3 columns of data called data.txt that looks like:

0.0000 1.0000 1.0
0.0634 1.0655 1.1353
0.1269 1.1353 1.28899916094
0.1903 1.2097 1.46345358199
0.2538 1.2889 1.6615188369
0.3173 1.3734 1.88639043926
...

I can plot this data, 2 versus 1 and 3 versus 1, very easily on the
same plot, with a legend, with log y values, and only for the xrange
between 2 and 3 with gnuplot:

set log y
set xrange[2:3]
plot 'data.txt' u 1:2 w l t 'apples', 'data.txt' u 1:3 w l t 'oranges'

Now, how do I do that same thing with matplotlob? Ie:

1. Both graphs overlayed on the same plot.
2. Semilogy. (log y values),
3. Only ploy for x in the range 2-3.
4. Legend for the two graphs on same plot.

Something like this:

import numpy as np
import matplotlib.pyplot as plt

x, apples, oranges = np.loadtxt('data.txt', unpack=True)
plt.semilogy(x, apples, label='apples')
plt.semilogy(x, oranges, label='oranges')
plt.legend()
plt.gca().set_xlim(2, 3)
plt.show()

There are many possible variations and styles. The basic point is to separate reading in the data from plotting it. Plotfile won't do what you want because it is designed to make separate subplots instead of plotting multiple lines on a single axes. Maybe doing the latter would be at least as useful, if not more, and could be enabled as an option with one more kwarg.

Eric

···

I have spent time looking through the documentation but I can't find
anyway to do this is any straightforward way. plotfile() looks
promising, but I can't seem to make it do the above. Thanks in
advance.

Joseph Smidt

Joseph_Smidt · April 29, 2009, 8:09am

Thanks everyone, this is exactly what I wanted.

···

--
------------------------------------------------------------------------
Joseph Smidt <josephsmidt@...287...>

Physics and Astronomy
4129 Frederick Reines Hall
Irvine, CA 92697-4575
Office: 949-824-3269

Matthias_Michler · April 29, 2009, 8:31am

Hello Eric, Hello list,

a year ago I also encountered the problem of "one file - one figure" of the
plotfile function. I would like to propose an addional functionality of using
one figure and several files in plotfile, because sometimes I don't want to
read data myself. I added a patch including the following changes:
- added a new keywordargument to plotfile 'use_cf': If use_cf isTrue plotfile
uses fig = gcf() instead of fig = figure() to suppress opening of a new
figure and therewith allowing to use the user preferred figure
- added a further new keyword argument 'names' to set x/ylabels in the case
there are no names in the csv-file

Furthermore I attached the modified plotfile_demo.py
(examples/pylab_examples/plotfile_demo.py) and some new data
(examples/data/data_x_x2_x3.csv).

Could this be useful?

Thanks in advance for any comments.

best regards
Matthias

plotfile_extended.patch (2.1 KB)

plotfile_demo.py (902 Bytes)

data_x_x2_x3.csv (132 Bytes)

···

On Wednesday 29 April 2009 09:20:17 Eric Firing wrote:

Joseph Smidt wrote:
> Okay, I am another gnuplot user trying to migrate over to matplotlib.
> I like what I see, but there are a couple things that are very easy to
> do in Gnuplot that I can't figure out how to do with matplotlib.
>
> I have a file with 3 columns of data called data.txt that looks like:
>
> 0.0000 1.0000 1.0
> 0.0634 1.0655 1.1353
> 0.1269 1.1353 1.28899916094
> 0.1903 1.2097 1.46345358199
> 0.2538 1.2889 1.6615188369
> 0.3173 1.3734 1.88639043926
> ...
>
> I can plot this data, 2 versus 1 and 3 versus 1, very easily on the
> same plot, with a legend, with log y values, and only for the xrange
> between 2 and 3 with gnuplot:
>
> set log y
> set xrange[2:3]
> plot 'data.txt' u 1:2 w l t 'apples', 'data.txt' u 1:3 w l t 'oranges'
>
> Now, how do I do that same thing with matplotlob? Ie:
>
> 1. Both graphs overlayed on the same plot.
> 2. Semilogy. (log y values),
> 3. Only ploy for x in the range 2-3.
> 4. Legend for the two graphs on same plot.

Something like this:

import numpy as np
import matplotlib.pyplot as plt

x, apples, oranges = np.loadtxt('data.txt', unpack=True)
plt.semilogy(x, apples, label='apples')
plt.semilogy(x, oranges, label='oranges')
plt.legend()
plt.gca().set_xlim(2, 3)
plt.show()

There are many possible variations and styles. The basic point is to
separate reading in the data from plotting it. Plotfile won't do what
you want because it is designed to make separate subplots instead of
plotting multiple lines on a single axes. Maybe doing the latter would
be at least as useful, if not more, and could be enabled as an option
with one more kwarg.

Eric

> I have spent time looking through the documentation but I can't find
> anyway to do this is any straightforward way. plotfile() looks
> promising, but I can't seem to make it do the above. Thanks in
> advance.
>
> Joseph Smidt

---------------------------------------------------------------------------
--- Register Now & Save for Velocity, the Web Performance & Operations
Conference from O'Reilly Media. Velocity features a full day of
expert-led, hands-on workshops and two days of sessions from industry
leaders in dedicated Performance & Operations tracks. Use code vel09scf
and Save an extra 15% before 5/3. http://p.sf.net/sfu/velocityconf
_______________________________________________
Matplotlib-users mailing list
Matplotlib-users@lists.sourceforge.net
matplotlib-users List Signup and Options

Eric_Firing1 · May 3, 2009, 12:17am

Matthias Michler wrote:

Hello Eric, Hello list,

a year ago I also encountered the problem of "one file - one figure" of the plotfile function. I would like to propose an addional functionality of using one figure and several files in plotfile, because sometimes I don't want to read data myself. I added a patch including the following changes:
- added a new keywordargument to plotfile 'use_cf': If use_cf isTrue plotfile uses fig = gcf() instead of fig = figure() to suppress opening of a new figure and therewith allowing to use the user preferred figure
- added a further new keyword argument 'names' to set x/ylabels in the case there are no names in the csv-file

Furthermore I attached the modified plotfile_demo.py (examples/pylab_examples/plotfile_demo.py) and some new data (examples/data/data_x_x2_x3.csv).

Could this be useful?

Thanks in advance for any comments.

Matthias,

I incorporated a slight modification of your changes (newfig=False instead of use_cf=True) together with changes I made to directly support what Joseph asked about. The result is in r7078.

I hesitated to make even these changes, though, because I think we should avoid trying to make plotfile into a do-all tool. It should be kept as something that may be handy for quick and dirty plotting in some situations; but when a user needs something beyond that, the better approach is for the user to simply use the pyplot or matplotlib API to achieve the desired result directly.

Eric

···

best regards
Matthias

On Wednesday 29 April 2009 09:20:17 Eric Firing wrote:

Joseph Smidt wrote:

Okay, I am another gnuplot user trying to migrate over to matplotlib.
I like what I see, but there are a couple things that are very easy to
do in Gnuplot that I can't figure out how to do with matplotlib.

I have a file with 3 columns of data called data.txt that looks like:

0.0000 1.0000 1.0
0.0634 1.0655 1.1353
0.1269 1.1353 1.28899916094
0.1903 1.2097 1.46345358199
0.2538 1.2889 1.6615188369
0.3173 1.3734 1.88639043926
...

I can plot this data, 2 versus 1 and 3 versus 1, very easily on the
same plot, with a legend, with log y values, and only for the xrange
between 2 and 3 with gnuplot:

set log y
set xrange[2:3]
plot 'data.txt' u 1:2 w l t 'apples', 'data.txt' u 1:3 w l t 'oranges'

Now, how do I do that same thing with matplotlob? Ie:

1. Both graphs overlayed on the same plot.
2. Semilogy. (log y values),
3. Only ploy for x in the range 2-3.
4. Legend for the two graphs on same plot.

Something like this:

import numpy as np
import matplotlib.pyplot as plt

x, apples, oranges = np.loadtxt('data.txt', unpack=True)
plt.semilogy(x, apples, label='apples')
plt.semilogy(x, oranges, label='oranges')
plt.legend()
plt.gca().set_xlim(2, 3)
plt.show()

There are many possible variations and styles. The basic point is to
separate reading in the data from plotting it. Plotfile won't do what
you want because it is designed to make separate subplots instead of
plotting multiple lines on a single axes. Maybe doing the latter would
be at least as useful, if not more, and could be enabled as an option
with one more kwarg.

Eric

I have spent time looking through the documentation but I can't find
anyway to do this is any straightforward way. plotfile() looks
promising, but I can't seem to make it do the above. Thanks in
advance.

Joseph Smidt

---------------------------------------------------------------------------
--- Register Now & Save for Velocity, the Web Performance & Operations
Conference from O'Reilly Media. Velocity features a full day of
expert-led, hands-on workshops and two days of sessions from industry
leaders in dedicated Performance & Operations tracks. Use code vel09scf
and Save an extra 15% before 5/3. http://p.sf.net/sfu/velocityconf
_______________________________________________
Matplotlib-users mailing list
Matplotlib-users@lists.sourceforge.net
matplotlib-users List Signup and Options