Develop Reference

Matplotlib shrink figure to size of axes, or otherwise match its aspect ratio

I have been using Cartopy to plot data using 'equal' aspect, resulting in all manner of non-square Axes sizes. These usually look OK in Jupyter notebooks, but when saving the images (or when doing more complicated operations like adding colorbars), the resulting Figures are often huge, with a lot of blank space around the Axes plotting area. They also look bad when using %matplotlib widget. An example is provided below.
It seems that the figure in this case is too big in at least one dimension. I would like to remove that extra space in the final output figure, without shrinking the size of the plotting area itself.
I know that I can adjust the figure size itself with .set_figwidth and .set_figheight, as well as setting figsize= upon creation. But I don't know how to figure out the correct dimensions to shrink the figure without shrinking the axes, and I haven't seen any way to do this automatically. What's the correct solution? I would like to avoid manually editing my images after creating them!

Jupyter: Seaborn pairplot difficult to set graph dimensions for?

I was trying to create a bivariate scatterplot of each variable against every other that are in a dataframe, and I found sns.pairplot() was exactly what I needed.
However, no matter what I do (and I have tried all of the advice found in this question), the plots keep coming out too spread out, as well as in general too big. In the picture below, only the first two rows and four columns out of 12 variables display on my entire screen.
I have found out that my use of
%config InlineBackend.figure_format = 'svg'
to create non-blurry graphs on my high-ppi screen is partially at blame, as without it, I instead get this graph, which fits perfectly on my screen, but is now too small and I would prefer to scroll slightly around while having a bigger pic.
(note: the additional options below have no effect)
How can I make the grid of plots customizable in its overall size as well as spacing? As it stands, no options work, and one graphics backend (the default one) produces too small graphs, while the 'svg' backend produces too large ones.
EDIT: Editing sns.set(rc={'figure.figsize':(x,y)}) or the height/ aspect options improve nothing: the former produces no change, while the latter two change how big the individual plots are (with height=1 making them indecipherable), but the overall "grid" is still as bulky and overly large as before.

Essentially you are asking how to display the figure in its original size in a jupyter notebook.
That translates into how to add scrollbars if it exceeds the room it's given by the layout of the output cell.
I think for the horizontal direction this can be done as follows. However for the vertical direction this does not seem to work.
%matplotlib inline
# Cell2
from IPython.display import display, HTML
CSS = """div.output_area img {max-width:None !important;max-height: None !important";}"""
display(HTML('<style>{}</style>'.format(CSS)))
# Cell3
import matplotlib.pyplot as plt
fig, ax = plt.subplots(ncols=8, figsize=(20,10))

Placing the legend outside the plot

I want to position the legend outside the drawing box. I do not find a clean way to do this. The main problem is having everything fit on the file saved. The only thing I have been able to figure out is this code:
#! /usr/bin/python
import matplotlib
# matplotlib.use('pdf')
from matplotlib.pyplot import *
subplot(111)
plot([1,2,3], label="test1")
l=legend(bbox_to_anchor=(1.05, 1), loc=2,borderaxespad=0)
tight_layout(rect=(0,0,0.8,1))
savefig('test.pdf')
There are a couple of caveats:
The tight_layout seems to be incompatible with matplotlib.use('pdf')
The 0.8 in the tight_layout has been found by trials and errors. If I replace label="test" by label="this is a very very long test", the legend will once again get out of the border of the file. I would like that to be adjusted automatically. I have not been able to retrieve the size of the legend. l.get_frame().get_width() seems to always return 1.0.
The tight_layout changes the size of the plot. What I would like to achieve is: specify the size of the plot in cm (or in inches). Placing the legend outside the plot. Having a file with the correct size in order that everything fits in, including the legend; without having to resort to trial and errors.

You can look at the answer to this question by Joe Kington.
The answer describes most of the options you can play around with regarding placement of legends.

Saving images in Python at a very high quality

How can I save Python plots at very high quality?
That is, when I keep zooming in on the object saved in a PDF file, why isn't there any blurring?
Also, what would be the best mode to save it in?
png, eps? Or some other? I can't do pdf, because there is a hidden number that happens that mess with Latexmk compilation.

If you are using Matplotlib and are trying to get good figures in a LaTeX document, save as an EPS. Specifically, try something like this after running the commands to plot the image:
plt.savefig('destination_path.eps', format='eps')
I have found that EPS files work best and the dpi parameter is what really makes them look good in a document.
To specify the orientation of the figure before saving, simply call the following before the plt.savefig call, but after creating the plot (assuming you have plotted using an axes with the name ax):
ax.view_init(elev=elevation_angle, azim=azimuthal_angle)
Where elevation_angle is a number (in degrees) specifying the polar angle (down from vertical z axis) and the azimuthal_angle specifies the azimuthal angle (around the z axis).
I find that it is easiest to determine these values by first plotting the image and then rotating it and watching the current values of the angles appear towards the bottom of the window just below the actual plot. Keep in mind that the x, y, z, positions appear by default, but they are replaced with the two angles when you start to click+drag+rotate the image.

Just to add my results, also using Matplotlib.
.eps made all my text bold and removed transparency. .svg gave me high-resolution pictures that actually looked like my graph.
import matplotlib.pyplot as plt
fig, ax = plt.subplots()
# Do the plot code
fig.savefig('myimage.svg', format='svg', dpi=1200)
I used 1200 dpi because a lot of scientific journals require images in 1200 / 600 / 300 dpi, depending on what the image is of. Convert to desired dpi and format in GIMP or Inkscape.
Obviously the dpi doesn't matter since .svg are vector graphics and have "infinite resolution".

You can save to a figure that is 1920x1080 (or 1080p) using:
fig = plt.figure(figsize=(19.20,10.80))
You can also go much higher or lower. The above solutions work well for printing, but these days you want the created image to go into a PNG/JPG or appear in a wide screen format.

Okay, I found spencerlyon2's answer working. However, in case anybody would find himself/herself not knowing what to do with that one line, I had to do it this way:
beingsaved = plt.figure()
# Some scatter plots
plt.scatter(X_1_x, X_1_y)
plt.scatter(X_2_x, X_2_y)
beingsaved.savefig('destination_path.eps', format='eps', dpi=1000)

In case you are working with seaborn plots, instead of Matplotlib, you can save a .png image like this:
Let's suppose you have a matrix object (either Pandas or NumPy), and you want to take a heatmap:
import seaborn as sb
image = sb.heatmap(matrix) # This gets you the heatmap
image.figure.savefig("C:/Your/Path/ ... /your_image.png") # This saves it
This code is compatible with the latest version of Seaborn. Other code around Stack Overflow worked only for previous versions.
Another way I like is this. I set the size of the next image as follows:
plt.subplots(figsize=(15,15))
And then later I plot the output in the console, from which I can copy-paste it where I want. (Since Seaborn is built on top of Matplotlib, there will not be any problem.)

python matplotlib blit to axes or sides of the figure?

I'm trying to refresh some plots that I have within a gui everytime I go once through a fitting procedure. Also, these plots are within a framw which can be resized, so the axes and labels etc need to be redrawn after the resizing. So was wondering if anyone knew how to update the sides of a figure using something like plot.figure.canvas.copy_from_bbox and blit. This appears to only copy and blit the background of the graphing area (where the lines are being drawn) and not to the sides of the graph or figure (where the labels and ticks are). I have been trying to get my graphs to update by trial and error and reading mpl documentation, but so far my code has jst become horrendously complex with things like self.this_plot.canvas_of_plot..etc.etc.. .plot.figure.canvas.copy_from_bbox... which is probably far too convoluted.
I know that my language might be a little off but I've been trying to read through the matplotlb documentation and the differences between Figure, canvas, graph, plot, figure.Figure, etc. are starting to evade me. So my first and foremost question would be:
1 - How do you update the ticks and labels around a matplotlib plot.
and secondly, since I would like to have a better grasp on what the answer to this question,
2 - What is the difference between a plot, figure, canvas, etc. in regards to the area which they cover in the GUI.
Thank you very much for the help.

All this can certainly be rather confusing at first!
To begin with, if you're chaining the ticks, etc, there isn't much point in using blitting. Blitting is just a way to avoid re-drawing everything if only some things are changing. If everything is changing, there's no point in using blitting. Just re-draw the plot.
Basically, you just want fig.canvas.draw() or plt.draw()
At any rate, to answer your first question, in most cases you won't need to update them manually. If you change the axis limits, they'll update themselves. You're running into problems because you're blitting just the inside of the axes instead of redrawing the plot.
As for your second question, a good, detailed overview is the Artist Tutorial of the Matplotlib User's Guide.
In a nutshell, there are two separate layers. One deals with grouping things into the parts that you'll worry about when plotting (e.g. the figure, axes, axis, lines, etc) and another that deals with rendering and drawing in general (the canvas and renderer).
Anything you can see in a matplotlib plot is an Artist. (E.g. text, a line, the axes, and even the figure itself.) An artist a) knows how to draw itself, and b) can contain other artists.
For an artist to draw itself, it uses the renderer (a backend-specific module that you'll almost never touch directly) to draw on a FigureCanvas a.k.a. "canvas" (an abstraction around either a vector-based page or a pixel buffer). To draw everything in a figure, you call canvas.draw().
Because artists can be groups of other artists, there's a hierarchy to things. Basically, something like this (obviously, this varies):
Figure
Axes (0-many) (An axes is basically a plot)
Axis (usually two) (x-axis and y-axis)
ticks
ticklabels
axis label
background patch
title, if present
anything you've plotted, e.g. Line2D's
Hopefully that makes things a touch clearer, anyway.
If you really do want to use blitting to update the tick labels, etc, you'll need to grab and restore the full region including them. This region is a bit tricky to get, because it isn't exactly known until after draw-time (rendering text in matplotlib is more complicated than rendering other things due to latex support, etc). You can do it, and I'll be glad to give an example if it's really what you want, but it's typically not going to yield a speed advantage over just drawing everything. (The exception is if you're only updating one subplot in a figure with lots of subplots.)