Merged streamline examples #8336

patniharshit · 2017-03-19T11:09:01Z

Based on discussion in #8155 a new pull request combining work with #8082.
Refs #7956

NelleV

This looks mostly good, but could you merge the figures into one using several subplots?

NelleV · 2017-03-19T19:11:48Z

examples/images_contours_and_fields/streamplot_demo.py

+Streamplot
+==========
+
+Demo of the `streamplot` function.


Can you remove this line? It is not informative.

patniharshit · 2017-03-20T07:09:07Z

I don't know why the one wth masking is behaving like this. Can you point out how to fix it?

WeatherGod · 2017-03-20T17:35:17Z

It is because of imshow() applying an equal aspect ratio.

patniharshit · 2017-03-21T06:53:33Z

Looks like

phobson · 2017-03-21T17:38:19Z

@patniharshit

I prefer the appearance of the plot with an equal aspect ratio. To get this in a desirable way, I'd used the gridspec module (it's always nice to show that off in other places)

fig = plt.figure(figsize=(7, 9))
gs = gspec.GridSpec(nrows=3, ncols=2, height_ratios=[1, 1, 2])
ax0 = fig.add_subplot(gs[0, 0])
ax1 = fig.add_subplot(gs[0, 1])
ax2 = fig.add_subplot(gs[1, 0])
ax3 = fig.add_subplot(gs[1, 1])
ax4 = fig.add_subplot(gs[2:, :])
ax4.set_aspect('equal')

patniharshit · 2017-03-23T14:54:26Z

QuLogic · 2017-03-24T23:07:42Z

examples/images_contours_and_fields/streamplot_demo.py

+fig.colorbar(strm.lines)
+ax3.set_title('Controlling Starting Points')
+
+# Displaying the starting points with red symbols.


They're blue, not red.

QuLogic · 2017-03-24T23:08:41Z

examples/images_contours_and_fields/streamplot_demo.py

+Y, X = np.mgrid[-w:w:100j, -w:w:100j]
+U = -1 - X**2 + Y
+V = 1 + X - Y**2
+speed = np.sqrt(U*U + V*V)


If I'm not mistaken, these are the same values as the ones at the top. Maybe give the ones for ax3 a different name, then you won't need to recreate these same values again here.

Oops, this comment is supposed to be about lines 62-65.

No, U,V are being changed at line 47-49.

afvincent

In overall the PR seems nice to me, thanks for working on this @patniharshit :)! I still have some comments and suggestion. Some are taken from #8082. Again, if you are a native English speaker and disagree with some of my remarks about phrasing: drop them ;).

FWIW, switching from cmap=plt.cm.name_of_a_cmap to cmap='name_of_a_cmap' is just a suggestion, but I think it may make the script a (very little) bit simpler.

If I am correct, you will also need to change the references in the docs that points to the deleted files. Looking at #8082, it should now be:

in doc/users/screenshots.rst , l. 74: .. plot:: mpl_examples/images_contours_and_fields/streamplot_demo.py
in doc/users/prev_whats_new/whats_new_1.2.rst , l. 155: .. plot:: mpl_examples/images_contours_and_fields/streamplot_demo.py

afvincent · 2017-03-25T10:49:14Z

examples/images_contours_and_fields/streamplot_demo.py

+Streamplot
+==========
+A streamplot, or streamline plot, is used to display 2D vector fields. This
+example shows a few features of the stream plot function:


stream plot <- streamplot?

afvincent · 2017-03-25T10:50:22Z

examples/images_contours_and_fields/streamplot_demo.py

+
+    * Varying the color along a streamline.
+    * Varying the density of streamlines.
+    * Varying the line width along a stream line.


stream line <- streamline?

afvincent · 2017-03-25T10:50:24Z

examples/images_contours_and_fields/streamplot_demo.py

+    * Varying the color along a streamline.
+    * Varying the density of streamlines.
+    * Varying the line width along a stream line.
+    * Controlling the start points of streamlines.


start points <- starting points?

afvincent · 2017-03-25T10:51:53Z

examples/images_contours_and_fields/streamplot_demo.py

+ax4.set_title('Streamline with Masking')
+
+ax4.imshow(~mask, extent=(-w, w, -w, w), alpha=0.5,
+          interpolation='nearest', cmap=plt.cm.gray, aspect='auto')


plt.cm.gray <- 'gray'?

afvincent · 2017-03-25T10:52:21Z

examples/images_contours_and_fields/streamplot_demo.py

+
+ax3 = fig.add_subplot(gs[1, 1])
+strm = ax3.streamplot(X, Y, U, V, color=U, linewidth=2,
+                     cmap=plt.cm.autumn, start_points=seed_points.T)


plt.cm.autumn <- 'autumn'?

afvincent · 2017-03-25T10:52:50Z

examples/images_contours_and_fields/streamplot_demo.py

+
+# Varying color along a streamline
+ax1 = fig.add_subplot(gs[0, 1])
+strm = ax1.streamplot(X, Y, U, V, color=U, linewidth=2, cmap=plt.cm.autumn)


plt.cm.autmn <- 'autumn'?

afvincent · 2017-03-25T11:01:22Z

examples/images_contours_and_fields/streamplot_demo.py

+# Controlling the starting points of the streamlines
+X, Y = (np.linspace(-3, 3, 100),
+        np.linspace(-3, 3, 100))
+U, V = np.mgrid[-3:3:100j, 0:0:100j]


Any reason for changing the U, V compared to the former streamplot_demo_start_points.py script? How bad would it be to simply re-use the same X, Y, U and V as for the other subplots? This way, you'd just need to create these (dummy) data at the beginning of the script (NB: it is already the case at ll. 18-22) and after that, the script is focused on the plotting aspect of streamplot. I think that it is also close to what @QuLogic had in mind in one of his remarks.

I have kept U, V exactly same as they are in previous streamplot_demo_start_points.py.

But if I comment out above lines and use same U,V everywhere then I get this plot -

Oh I think that I understand why the current situation was weird for both of us!

The current example in the official docs is indeed the one that you are refering to. However, the starting point example was changed in #6672 (by myself, which is why I was having the feeling that you changed U and V here) to use the same data as the other streamplot examples, which I think is really more demonstrative to show the different plotting options of streamplot. The related version of the example is only in our devdocs gallery (I guess because the PR was milestoned for 2.1 and not backported into 2.0). So I would definitively support to use the same data (built once for all at the top of the script) for all the subplots :).

afvincent · 2017-03-25T11:04:07Z

examples/images_contours_and_fields/streamplot_demo.py

+Y, X = np.mgrid[-w:w:100j, -w:w:100j]
+U = -1 - X**2 + Y
+V = 1 + X - Y**2
+speed = np.sqrt(U*U + V*V)


If one decides to always plot the same data among all the different subplots, the 5 lines above could be deleted.

Huh, why is this comment here?! This is the opposite of what I had in mind ^^... Actually lines 19-23 would be the only place where I would suggest to create all the dummy data!

afvincent · 2017-03-25T11:14:31Z

examples/images_contours_and_fields/streamplot_demo.py

+==========
+Streamplot
+==========
+A streamplot, or streamline plot, is used to display 2D vector fields. This


I am not sure that it is a mandatory requirement, but you may want to add a blank line betwenn the second ===== line and the first line of the paragraph. Maybe @NelleV would be able to confirm what SG and rST really require :).

I don't think it is.

Then one remaining argument in favor of adding a blank line would be consistency among our example scripts, as a lot of other examples in the docs seem to have one such blank line ;). But I am also fine with no blank line if SG and rST are OK!

Blank line added to maintain consistency among all examples.

need to re-review

afvincent · 2017-03-25T16:23:23Z

examples/images_contours_and_fields/streamplot_demo.py

+# Displaying the starting points with blue symbols.
+ax3.plot(seed_points[0], seed_points[1], 'bo')
+
+ax3.axis((-3, 3, -3, 3))


I forgot to comment on this one during my review, sorry. Is this line really needed (I do not see the equivalent for the previous subplots)? If indeed it is and one actually uses the same dummy data accross all the subplots, I would then rather suggest using ax3.axis((-w, w, -w, w)) instead (if I understand correctly the purpose of this line).

Other subplots were going from -3 to 3 but this one was not, so probably this line was put there to have uniformity in all plots. I have kept this. But it can be deleted considering this is supposed to be an example, thus it should use minimal code to achieve target.

You are right. Due to the reduced amount of streamlines that are plotted, it seems like the axis is not exactly the same as for the other plots. So it might be wiser to keep it :). 👍 for the new version ax3.axis((-w, w, -w, w)).

afvincent

A few remaining comments, among which my biggest one will be to use the same dummy data (defined at the beginning of the script) for all the subplots. But apart from that, the PR starts to look really good to me :).

afvincent · 2017-03-25T22:56:48Z

examples/images_contours_and_fields/streamplot_demo.py

+Y, X = np.mgrid[-w:w:100j, -w:w:100j]
+U = -1 - X**2 + Y
+V = 1 + X - Y**2
+speed = np.sqrt(U*U + V*V)


Huh, why is this comment here?! This is the opposite of what I had in mind ^^... Actually lines 19-23 would be the only place where I would suggest to create all the dummy data!

afvincent · 2017-03-25T22:57:24Z

examples/images_contours_and_fields/streamplot_demo.py

+ax1 = fig.add_subplot(gs[0, 1])
+strm = ax1.streamplot(X, Y, U, V, color=U, linewidth=2, cmap='autumn')
+fig.colorbar(strm.lines)
+ax1.set_title('Varying color')


Varying Color?

afvincent · 2017-03-25T22:59:00Z

examples/images_contours_and_fields/streamplot_demo.py

+
+ax4 = fig.add_subplot(gs[2:, :])
+ax4.streamplot(X, Y, U, V, color='r')
+ax4.set_title('Streamline with Masking')


Plural form (as there are several streamlines), or "Streamplot with Masking"?

Changed to "Streamplot with masking".

afvincent · 2017-03-25T23:08:15Z

examples/images_contours_and_fields/streamplot_demo.py

+==========
+
+A stream plot, or stream line plot, is used to display 2D vector fields. This
+example shows a few features of the stream plot function:


Again, when it comes to English, I may be totally wrong, but I think "the streamplot function" may be more correct than the current "the stream plot function", as it is refering to the streamplot function and not to a stream plot. I guess I was not clear enough: it was actually the meaning of my previous comment ^^. However, for the other utterances of "stream plot"/"stream line," I will let a native English speaker (who may be you!) decide if it is better than "streamplot"/"streamline" or not.

I think generally, "streamline" is a word and "streamplot" is not, unless you're referring to the function itself (as in this case.)

QuLogic · 2017-03-25T23:43:55Z

@NelleV Should this be renamed as plot_something.py?

QuLogic · 2017-03-25T23:45:03Z

I don't think example deletion should be backported, so I've upped the milestone.

afvincent · 2017-04-02T08:31:04Z

Travis failures seem unrelated (they are about test_mixedsubplots…)

afvincent

Last tiny nitpicking, based on @QuLogic comment about writing “streamline” with a single word (basically it reverts some of the changes made by a530e79).

Besides, pinging @NelleV about the other @QuLogic's comment: should this be renamed as plot_something.py (for Sphinx gallery sake IIUC)?

Apart from that, the new version of the example script looks fine, thanks @patniharshit :)!

Question to core devs who may pass by: if @patniharshit is comfortable with git, would it be relevant to rebase, and squash a bit the commits?

afvincent · 2017-04-02T08:44:53Z

examples/images_contours_and_fields/streamplot_demo.py

+Streamplot
+==========
+
+A stream plot, or stream line plot, is used to display 2D vector fields. This


afvincent · 2017-04-02T08:45:06Z

examples/images_contours_and_fields/streamplot_demo.py

+A stream plot, or stream line plot, is used to display 2D vector fields. This
+example shows a few features of the streamplot function:
+
+    * Varying the color along a stream line.


afvincent · 2017-04-02T08:45:18Z

examples/images_contours_and_fields/streamplot_demo.py

+example shows a few features of the streamplot function:
+
+    * Varying the color along a stream line.
+    * Varying the density of stream lines.


streamlines

afvincent · 2017-04-02T08:45:26Z

examples/images_contours_and_fields/streamplot_demo.py

+
+    * Varying the color along a stream line.
+    * Varying the density of stream lines.
+    * Varying the line width along a stream line.


afvincent · 2017-04-02T08:45:33Z

examples/images_contours_and_fields/streamplot_demo.py

+    * Varying the color along a stream line.
+    * Varying the density of stream lines.
+    * Varying the line width along a stream line.
+    * Controlling the starting points of stream lines.


streamlines

afvincent · 2017-04-02T08:45:43Z

examples/images_contours_and_fields/streamplot_demo.py

+    * Varying the density of stream lines.
+    * Varying the line width along a stream line.
+    * Controlling the starting points of stream lines.
+    * Stream lines skipping masked regions and NaN values.


Streamlines

afvincent

LGTM :).

A few questions to other devs:

should this be renamed as “plot_something.py” for Sphinx gallery sake? (@NelleV , pinging our SG expert if she is around)
should it be rebased and some of the commits squashed if @patniharshit is comfortable with git?

phobson · 2017-04-05T15:26:27Z

@afvincent my answer is "yes" to both of those.

patniharshit · 2017-04-05T20:25:33Z

Renamed, rebased and commits squashed.

afvincent

Minor fixes needed. Apart from that, the other Travis failures seems to be spurious ones.

afvincent · 2017-04-10T07:44:32Z

doc/users/prev_whats_new/whats_new_1.2.rst

@@ -152,7 +152,7 @@ In addition to simply plotting the streamlines of the vector field,
 line widths of the streamlines to a separate parameter, such as the speed or
 local intensity of the vector field.

-.. plot:: mpl_examples/images_contours_and_fields/streamplot_demo_features.py
+.. plot:: mpl_examples/images_contours_and_fields/streamplot_demo.py


'streamplot_demo.py' should be changed into 'plot_streamplot.py', shouldn't it?

afvincent · 2017-04-10T07:44:44Z

doc/users/screenshots.rst

@@ -71,7 +71,7 @@ a vector field. In addition to simply plotting the streamlines, it allows you
 to map the colors and/or line widths of streamlines to a separate parameter,
 such as the speed or local intensity of the vector field.

-.. plot:: mpl_examples/images_contours_and_fields/streamplot_demo_features.py
+.. plot:: mpl_examples/images_contours_and_fields/streamplot_demo.py


Here too, 'streamplot_demo.py' should be changed into 'plot_streamplot.py', shouldn't it?

patniharshit · 2017-04-10T09:09:17Z

Oh forgot this. Will do it asap.

afvincent · 2017-04-11T09:19:49Z

LGTM. Thanks @patniharshit :) !

QuLogic · 2017-04-11T23:04:03Z

Ugh, I was going to merge this, but it looks like it needs a rebase with that sphinx-gallery PR merged.

dstansby · 2017-04-12T10:46:32Z

I'm going to attempt a rebase and merge

files

dstansby · 2017-04-12T11:09:16Z

Okay, that looks like it worked, but since it's the first time I've done a rebase and force-push I'll wait for CI to pass before merging.

SKPsanjeevi · 2017-04-28T18:05:17Z

I see that the current version of the plot_streamplot.py from repo is giving errors/not executing.

QuLogic · 2017-04-28T21:32:16Z

I see no errors in the latest CI build; note that examples from master are not guaranteed to work on any released version.

SKPsanjeevi · 2017-04-28T21:36:13Z

How can I also replicate it? I am spending time on fixing the error from master. Could you please let me know, how to perform such a check?

QuLogic · 2017-04-28T21:37:30Z

Replicate what? There is no error in master.

SKPsanjeevi · 2017-04-28T21:39:05Z

I get the following error:

Traceback (most recent call last):
  File "plot_streamplot.py", line 50, in <module>
    cmap='autumn', start_points=seed_points.T)
  File "/usr/lib/python2.7/dist-packages/matplotlib/__init__.py", line 1814, in inner
    return func(ax, *args, **kwargs)
  File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_axes.py", line 4464, in streamplot
    zorder=zorder)
  File "/usr/lib/python2.7/dist-packages/matplotlib/streamplot.py", line 140, in streamplot
    sp2[:, 0] += np.abs(x[0])
ValueError: operands could not be broadcast together with shapes (6,) (100,) (6,)

SKPsanjeevi · 2017-04-28T21:45:02Z

The syntax for streamline says that x and y should be 1D array and u and v as 2D array. However in the example, both x and y are made as 2D grids using np.mgrid. When they are made just as 1D array, the issue is fixed.

phobson · 2017-04-28T21:47:41Z

@pssk1988 -- are you working with a matplotlib installed from the latest github master?

SKPsanjeevi · 2017-04-28T21:54:00Z

@phobson : I don't think so :/ I will look for using the github matplotlib.

SKPsanjeevi · 2017-04-28T22:00:24Z

I am really sorry to bother you all without using the github version of matplotlib. Now it works fine.

afvincent added the Documentation label Mar 19, 2017

NelleV previously requested changes Mar 19, 2017

View reviewed changes

dstansby added this to the 2.0.1 (next bug fix release) milestone Mar 24, 2017

QuLogic reviewed Mar 24, 2017

View reviewed changes

QuLogic mentioned this pull request Mar 25, 2017

merge streamplot_demo_features streamplot_demo_masking streamplot_demo_start_points #8155

Closed

afvincent requested changes Mar 25, 2017

View reviewed changes

afvincent mentioned this pull request Mar 25, 2017

Merged and improved the streamplot demonstration #8082

Closed

afvincent reviewed Mar 25, 2017

View reviewed changes

afvincent requested changes Mar 25, 2017

View reviewed changes

QuLogic modified the milestones: 2.1 (next point release), 2.0.1 (next bug fix release) Mar 25, 2017

anntzer mentioned this pull request Mar 25, 2017

Suggested improvements to the gallery #7956

Closed

16 tasks

afvincent requested changes Apr 2, 2017

View reviewed changes

afvincent approved these changes Apr 5, 2017

View reviewed changes

afvincent changed the title ~~Merged streamline examples~~ [MRG+1] Merged streamline examples Apr 5, 2017

patniharshit force-pushed the mergedStreamline branch from 3302135 to 7baa01e Compare April 5, 2017 20:24

afvincent requested changes Apr 10, 2017

View reviewed changes

afvincent approved these changes Apr 11, 2017

View reviewed changes

patniharshit added 4 commits April 12, 2017 11:49

Merge all figure in streamplot examples in one plot

bf2b43b

Add gridspec to have equal aspect ratio and remove references of deleted

92b1196

files

Remove data regeneration

913fdb7

Update example name in docs

d56b5ce

dstansby force-pushed the mergedStreamline branch from fe5dc72 to d56b5ce Compare April 12, 2017 11:08

dstansby self-assigned this Apr 12, 2017

dstansby merged commit b81576c into matplotlib:master Apr 12, 2017

dstansby changed the title ~~[MRG+1] Merged streamline examples~~ Merged streamline examples Apr 12, 2017

patniharshit deleted the mergedStreamline branch April 12, 2017 11:59

dstansby removed their assignment May 27, 2017

Uh oh!

Merged streamline examples #8336

Merged streamline examples #8336

Uh oh!

Conversation

patniharshit commented Mar 19, 2017

Uh oh!

NelleV left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

patniharshit commented Mar 20, 2017

Uh oh!

WeatherGod commented Mar 20, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

patniharshit commented Mar 21, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

phobson commented Mar 21, 2017

Uh oh!

patniharshit commented Mar 23, 2017

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

afvincent left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

patniharshit Mar 25, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

WeatherGod commented Mar 20, 2017 •

edited

Loading

patniharshit commented Mar 21, 2017 •

edited

Loading

patniharshit Mar 25, 2017 •

edited

Loading