mpl_toolkits axes & grid

Community matplotlib-users
1

Hi,

It is quite some time ago since I was intensively using MPL. Perhaps I will not get the terminology right anymore. And now I hope to get some pointer for this particular problem:

I would like to arange some line plots using a grid from left to right in a single plot / figure instance. Each of these plots is created from vectors (numpy arrays) of a different length. If I would place subplots of equal size to place these individual plots from left to right, all will appear with a different “stretch/squeeze”.

My question now is: Is there a way to place different subplots of different sizes automatically? For all my plots I will encounter different number of individual line plots of different size.

My approach so far (pseudocode):

from mpl_toolkits.axes_grid1 import Grid
import pylab as plt

fig = plt.figure(1, (5.5, 3.5))
grid = Grid(fig, 111, # similar to subplot(111)
nrows_ncols = (1, number_of_subplots),
axes_pad = 0.1,
add_all=True,
label_mode = “L”,
)

for index in xrange(number_of_subplots):

grid[index].???

And now I just do not know how to get any further? How will I get the necessary extension for each subfigure/plot? Which is the most efficient way to accomplish all this? Should I import a different class, other than Grid?

TIA

Regards,

Christian

2

So, are you wanting them to all have the same x and y coordinates, or maybe you want the aspect ratio to be the same?

I am not exactly quite sure what you mean by “extension”.

Ben Root

···

On Fri, Mar 15, 2013 at 12:41 PM, Christian Meesters <meesters@…380…> wrote:

Hi,

It is quite some time ago since I was intensively using MPL. Perhaps I will not get the terminology right anymore. And now I hope to get some pointer for this particular problem:

I would like to arange some line plots using a grid from left to right in a single plot / figure instance. Each of these plots is created from vectors (numpy arrays) of a different length. If I would place subplots of equal size to place these individual plots from left to right, all will appear with a different “stretch/squeeze”.

My question now is: Is there a way to place different subplots of different sizes automatically? For all my plots I will encounter different number of individual line plots of different size.

My approach so far (pseudocode):

from mpl_toolkits.axes_grid1 import Grid
import pylab as plt

fig = plt.figure(1, (5.5, 3.5))
grid = Grid(fig, 111, # similar to subplot(111)

             nrows_ncols = (1, number_of_subplots),
             axes_pad = 0.1,
             add_all=True,
             label_mode = "L",
             )

for index in xrange(number_of_subplots):

grid[index].???

And now I just do not know how to get any further? How will I get the necessary extension for each subfigure/plot? Which is the most efficient way to accomplish all this? Should I import a different class, other than Grid?

3

Hi David and Ben and everybody reading along,

Apparently I did not phrase my question too well.

To clarify:

I would like to have some 'subplots' with a width proportional to
the amount of (equally spaced) data points.

It seems to me that gridspec (http://matplotlib.org/users/gridspec.html)
is just the tool I need - with this I only need to calculate the
relative sizes of my subplots, but that's easy. Alas, I did not discover
it until tonight.

Thanks a lot for your help!
Christian