switched code in example to:

data = [('apples', 2), ('oranges', 3), ('peaches', 1)]
fruit, value = zip(*data)

fig, ax = plt.subplots()
ax.bar(fruit, value, align='center', color='gray')
plt.show()

This example is inspired from my severe usage of MATLAB overlay plotting where I plot on a figure and based on its distribution/look I do some operation in backend like moving the image to another directory etc. Hoping this example would become handy for someone like me(who moved from MATLAB plotting)
Discussion : Twitter Link

• /mpl-data/sample_data/axes_grid folder appears to no longer exists as of matplotlib v2.2.2
• added /assets folder in repo containing dependent numpy pickle, 'bivariate_normal.npy' file
• revised load of data in AnatomyOfMatplotlib-Part2-Plotting_Methods_Overview.ipynb to reflect this change
• tested successfully with matplotlib v2.2.2 and python v3.6.6

Added geometry to the Ugly Tie polygon to look like a knot.

It remains a single polygon so color will affect both visible parts.

At large zooms/resolutions a connection between the right side of the knot and the main tie is visible because the points on right side of the knot are not perfectly in-line with the upper right corner of the tie where the two larger parts of the shape are visible. If this becomes an issue, doing some math to find evenly dividing, aligned points near the current values would make the connecting section of the polygon zero width.

I realize this is a bit last-minute and a big change, but I really feel like we were missing a good overview of the various plotting methods.

I've added a new Part 2 (and renamed the other sections) to cover this: http://nbviewer.ipython.org/url/geology.beer/scipy2015/tutorial/AnatomyOfMatplotlib-Part2-Plotting_Methods_Overview.ipynb

I've tried to make a lot of nice summary images of the most commonly-used plotting methods. The "full" gallery can be very overwhelming, so it's useful to give people a condensed version. Also, these

80% of the new Part 2 is just quickly looking those images so that people are vaguely aware of what's out there. The code to generate them is also there to serve as an example.

The new section only goes over bar, fill_between and imshow in more detail. It's not anywhere near as long as it looks at first glance.

The ipython notebook format has changed slightly in recent versions. Notebooks in the old format are automatically converted when they're opened with a more recent version, but I wanted to go ahead and commit the new format versions.

Otherwise the diffs will be very difficult to read.

I also added a .gitignore to ignore the hidden checkpoint folder IPython adds, if ipython notebook is run from the source directory.

Hello matplotlib developers,

I was watching the Youtube recording: Anatomy of Matplotlib from SciPy 2018, and I have a question about AnatomyOfMatplotlib/solutions/2.2-vmin_vmax_imshow_and_colorbars.py

From line 17 to 18...

for ax, data in zip(axes, [data1, data2, data3]): im = ax.imshow(data, vmin=0, vmax=3, interpolation='nearest')

I am assuming data3 has bigger values, followed by data2 and data2, since data3 is multiplied by 3. Suppose if I switch the order of the list in line 17 from:

for ax, data in zip(axes, [data1, data2, data3]):

to:

for ax, data in zip(axes, [data3, data2, data1]):

So, the last im object would data1 which has a 10 by 10 array with max value of 1. Since we are giving the last im object to make the colorbar, would that mean the range color bar spans from 0 to around 1? Or does matplotlib somehow manage to look at all three plotted imshows and perceive that the maximum value amongst the three imshows is around 3?

Thank you!

Seems like the example data used in chapter 2 at the colorbar example is no longer supported as of py 3.1. bivariate_normal.npy is not in any folder and has apparantly been discontinued.

First off, IPython/Jupyter has recently had a .ipynb format change, so these diffs are rather messy. If I'd thought about it more, I would have made that a separate commit, but I didn't realize until edits were underway.

At any rate, I've changed Part1 rather significantly. I pruned some things out and expanded others. I'm intending to add another section detailing basic categories of plotting functions, so I removed several of the references to those in this section.

Even after these changes, Part1 is still rather long. I might split it (particularly the part after the second exercise and before the third) into another section.

At any rate hopefully you can see where I'm going with this. Thanks, and looking forward to teaching here in a few weeks!

Just finished going through the notebooks with Python3 and everything worked fine except for having to manually modify all the print statements. Figured it could be made to seamlessly work with both Python 2 and 3 by simply using a from __future__ import print_function.

Also threw out some extraneous sentences. Keep things simple and straight-forward. Resist the temptation to reveal everything at once. I will leave this up for a little bit for comment and then merge later today.

Backend is no longer necessary IMO. Using a backend results in the following error on Jupyter.

Javascript Error: IPython is not defined

Also adding %matplotlib inline before importing matplotlib resolves the displaying of graphs.

Should I fix them in the notebooks and send a PR?

Thanks.

In part 2, the example is too much to do at once. Rather, it would make sense to build up that example as more is taught. Perhaps a new feature for IPython notebooks would be useful (floating cells?)