[MRG + 1] 10 more examples fixed with matplotlib 2.0 updates #9030

thismlguy · 2017-06-07T04:48:13Z

10 more examples updated with matplotlib 2.0 updates.

Fixes issue #8364

The old and new plots will be posted soon after circleCI build completes.

…pdate

…ep8 fix

…2.0 update

thismlguy · 2017-06-07T17:40:49Z

calibration/plot_calibration.html

Old:

New:

Second image unchanged

thismlguy · 2017-06-07T17:41:45Z

classification/plot_lda_qda.html

Old:

New:

thismlguy · 2017-06-07T17:43:16Z

cluster/plot_agglomerative_clustering.html

Old:

New:

thismlguy · 2017-06-07T17:44:08Z

cluster/plot_birch_vs_minibatchkmeans.html

Old:

New:

thismlguy · 2017-06-07T17:46:29Z

cluster/plot_cluster_iris.html

Old:

New:

thismlguy · 2017-06-07T17:47:10Z

cluster/plot_kmeans_assumptions.html

Old:

New:

thismlguy · 2017-06-07T17:49:07Z

cluster/plot_kmeans_silhouette_analysis.html

Old:

New:

thismlguy · 2017-06-07T17:51:04Z

cluster/plot_ward_structured_vs_unstructured.html

Old:

New:

thismlguy · 2017-06-07T17:52:24Z

covariance/plot_outlier_detection.html

Old:

New:

thismlguy · 2017-06-07T17:53:57Z

plot_johnson_lindenstrauss_bound.html

Old:

New:

Other images unchanged.

thismlguy · 2017-06-07T17:57:14Z

added the old-new plots for the current circleCI build.

agramfort · 2017-06-07T19:45:02Z

besides plot_voting_decision_regions new looks always better

thismlguy · 2017-06-08T02:38:49Z

@agramfort - i'm sorry could you point me to the example which you are referring to?

agramfort · 2017-06-08T11:35:04Z

forget it for my previous comment. But plot_agglomerative_clustering and plot_kmeans_assumptions look better before. When dots are small putting a black border makes everything almost black.

…ustering

thismlguy · 2017-06-08T17:41:51Z

@agramfort - i found the new ones a bit cleaner. But I see you point. I've pushed a new commit with changes. i'm good with both. i'll update the plots here once the builds are complete.

thismlguy · 2017-06-09T01:44:08Z

CircleCI is giving this error - "Your build output is too large to display in the browser. Only the first 400000 characters are displayed." I tried searching couldn't find much.

@agramfort, @amueller - any thoughts for the reason?

agramfort · 2017-06-09T07:35:51Z

here is the link:

https://circleci.com/api/v1.1/project/github/scikit-learn/scikit-learn/10969/output/11/0?file=true

amueller · 2017-06-19T16:18:15Z

LGTM

amueller · 2017-06-19T16:19:09Z

examples/cluster/plot_cluster_iris.py

-    ax = Axes3D(fig, rect=[0, 0, .95, 1], elev=48, azim=134)
-
-    plt.cla()
+    ax = plt.subplot(2, 2, fignum, projection='3d',


which matplotlib version does this require?

actually 1.5 requires this. Issue was that I wasn't importing Axes3D because it wasn't being used directly. But I found later that this projection=3d requires Axes3D imported. So it worked after adding that import. All good now.

raghavrv · 2017-06-20T13:39:04Z

Can you fix the conflicts? This would be nice to have in 0.19

raghavrv · 2017-06-20T13:40:40Z

Never mind I fixed it myself ;) Let's wait for CIs. Meanwhile @amueller can we merge it?

amueller · 2017-06-20T14:49:30Z

@raghavrv if it has your +1 you can merge, circle and travis are green

raghavrv · 2017-06-21T14:05:05Z

Thanks a lot @aarshayj :)

thismlguy · 2017-06-21T16:19:59Z

thanks @raghavrv! I'll take care of the CIs on my end in future PRs :)

Use 4 plots instead of one so that unsupervised_learning.rst can show png separately. Partially reverts #9030. Minor changes for consistent labeling.

…learn#9030) * examples/covariance/plot_outlier_detection.py - matplotlib2.0 update * examples/cluster/plot_kmeans_silhouette_analysis.py - matplotlib2.0 update * examples/cluster/plot_birch_vs_minibatchkmeans.py - matplotlib2.0 + pep8 fix * examples/cluster/plot_cluster_iris.py - matplotlib2.0 update * examples/cluster/plot_agglomerative_clustering.py - matplotlib2.0 update * examples/cluster/plot_ward_structured_vs_unstructured.py - matplotlib2.0 update * examples/cluster/plot_kmeans_assumptions.py - matplotlib2.0 update * examples/classification/plot_lda_qda.py - matplotlib2.0 + pep8 fix * examples/calibration/plot_calibration.py - matplotlib2.0 update * examples/plot_johnson_lindenstrauss_bound.py - matplotlib2.0 update * flake8 changes * reversing changes for plot_kmeans_assumptions & plot_agglomerative_clustering * cluster/plot_cluster_iris.py - major fixes * examples/cluster/plot_cluster_iris.py - flake8 fix * examples/cluster/plot_cluster_iris.py - 3d projection error fix * cluster/plot_cluster_iris.py - elevation and azimuth setting

Use 4 plots instead of one so that unsupervised_learning.rst can show png separately. Partially reverts scikit-learn#9030. Minor changes for consistent labeling.