grassmannian test visuaiization for 2D #1803

christine051601 · 2023-01-28T01:58:27Z

test for 2d visualization

review-notebook-app · 2023-01-28T01:58:31Z

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

codecov · 2023-01-28T21:03:36Z

Codecov Report

Merging #1803 (d7596ce) into master (6f37453) will decrease coverage by 0.85%.
The diff coverage is 38.67%.

@@            Coverage Diff             @@
##           master    #1803      +/-   ##
==========================================
- Coverage   91.50%   90.64%   -0.85%     
==========================================
  Files         136      137       +1     
  Lines       13738    13963     +225     
==========================================
+ Hits        12569    12655      +86     
- Misses       1169     1308     +139

Flag	Coverage Δ
autograd	`86.59% <38.67%> (-0.82%)`	⬇️
numpy	`86.22% <38.67%> (-0.87%)`	⬇️
pytorch	`79.79% <38.67%> (-0.72%)`	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
geomstats/visualization/grassmannian.py	`38.67% <38.67%> (ø)`
geomstats/geometry/positive_reals.py	`93.48% <0.00%> (-2.17%)`	⬇️
geomstats/geometry/stratified/wald_space.py	`88.94% <0.00%> (-1.27%)`	⬇️
geomstats/_backend/pytorch/__init__.py	`96.63% <0.00%> (-0.24%)`	⬇️
geomstats/learning/expectation_maximization.py	`82.69% <0.00%> (+2.24%)`	⬆️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

…ChristineWan

ninamiolane

Address my code review comments, specifically the ones on respecting the hw's guidelines, and resubmit.

ninamiolane · 2023-02-13T16:35:51Z

geomstats/visualization/grassmannian.py

+    Takes Grassmannian generated projection matrix
+    and converts it to a plotable 2D point.
+
+    INPUT: Array of 2x2 projection matricies


Use numpy's convention for docstrings, see guidelines:
https://github.com/bioshape-lab/ece594n/tree/main/hw_geomviz

ninamiolane · 2023-02-13T16:36:27Z

geomstats/visualization/grassmannian.py

+    Takes Grassmannian generated projection matrix and
+    converts it to a plotable 3D point.
+
+    INPUT: Array of 3x3 projection matricies or a single projection matrix


Follow guidelines for docstrings: https://github.com/bioshape-lab/ece594n/tree/main/hw_geomviz

ninamiolane · 2023-02-13T16:36:31Z

geomstats/visualization/grassmannian.py

+def two_d_to_projection(vector):
+    """Take 2D point on manifold and gets its corresponding projection.
+
+    INPUT: Point on manifold as a 2x1 vector


Follow guidelines for docstrings: https://github.com/bioshape-lab/ece594n/tree/main/hw_geomviz

ninamiolane · 2023-02-13T16:36:59Z

geomstats/visualization/grassmannian.py

+    return projection
+
+
+def submersion(point, k):


Remove this function entirely (it is already coded in geomstats)

ninamiolane · 2023-02-13T16:37:30Z

geomstats/visualization/grassmannian.py

+    return gs.concatenate([row_1, row_2], axis=-2)
+
+
+def _squared_d_g_a(point_a, point_b, metric):


name this function _squared_dist_grad_point_a for consistency with next funciton

ninamiolane · 2023-02-13T16:39:29Z

geomstats/visualization/grassmannian.py

+            ax.set_zticks([])
+            plt.show()
+
+    def plot_tangent_space(self):


Keep this method! But follow guidelines for docstring

ninamiolane · 2023-02-13T16:39:39Z

geomstats/visualization/grassmannian.py

+    def plot_tangent_space(self):
+        """Plot Grassmanian tangent space.
+
+        Warning: Under construction, do not use.


Explain which case works in details

ninamiolane · 2023-02-13T16:40:16Z

geomstats/visualization/grassmannian.py

+        plt.plot(x_part, y_part, "-g")
+
+
+class GrassmannianCanonicalMetric(MatricesMetric, RiemannianMetric):


Remove this class entirely. If you need the metric, import it direclty from geomstats:

from geomstats.geometry.grassmannian import GrassmannianCanonicalMetric

ninamiolane · 2023-02-13T16:41:09Z

tests/tests_geomstats/test_visualization_grassmannian.py

+        """Test the plotting function of Grassmannian for 2 dimensions."""
+        self.grassmannian21.plot(True)
+
+    def test_plot_2d_render(self):


Follow the hw's guidelines: the name of the test function should be the same as the name of the function being tests

i.e. rename to test_plot_rendering_2d

ninamiolane · 2023-02-13T16:41:40Z

geomstats/visualization/grassmannian.py

+The Grassmannian :math:`Gr(n, k)` is the manifold of k-dimensional
+subspaces in n-dimensional Euclidean space.
+
+Lead author: Olivier Peltre.


Remove: the lead author of this file is not Olivier Peltre

review-notebook-app · 2023-02-13T16:47:18Z

View / edit / reply to this conversation on ReviewNB

ninamiolane commented on 2023-02-13T16:47:17Z
----------------------------------------------------------------

change the name of the ipynb file so that there is no uppercase (everything should be lower case)

review-notebook-app · 2023-02-13T16:47:19Z

View / edit / reply to this conversation on ReviewNB

ninamiolane commented on 2023-02-13T16:47:18Z
----------------------------------------------------------------

Remove that sentence "The Grassmanian can be expressed in various ways: a set, a differentiable manifold, a set of orthogonal projections, a homogeneous space, or a scheme. One of the most significant qualities of a Grassmannian manifold is that they are an orbit space of the Steifel Manifold v

n

,

k

��,�

of orthonormal k

�

frames in G

n

��

. (Wolfram)"

review-notebook-app · 2023-02-13T16:47:20Z

View / edit / reply to this conversation on ReviewNB

ninamiolane commented on 2023-02-13T16:47:20Z
----------------------------------------------------------------

Remove this image

review-notebook-app · 2023-02-13T16:47:22Z

View / edit / reply to this conversation on ReviewNB

ninamiolane commented on 2023-02-13T16:47:21Z
----------------------------------------------------------------

Remove the sentence: "Today, Grassmanian manifolds are applied to computer vision problems such as video-based shape recognition. In computer vision applications, Grassman manifolds are used to compute distances between different images and shapes. "

review-notebook-app · 2023-02-13T16:47:23Z

View / edit / reply to this conversation on ReviewNB

ninamiolane commented on 2023-02-13T16:47:22Z
----------------------------------------------------------------

Remove that image and the text below.

review-notebook-app · 2023-02-13T16:47:23Z

View / edit / reply to this conversation on ReviewNB

ninamiolane commented on 2023-02-13T16:47:23Z
----------------------------------------------------------------

Remove this section III, as well as the references it cites (from section on references)

review-notebook-app · 2023-02-13T16:47:25Z

View / edit / reply to this conversation on ReviewNB

ninamiolane commented on 2023-02-13T16:47:25Z
----------------------------------------------------------------

Line #8.    # TODO: Change these paths with the corresponding paths on your computer

Remove these lines and the TODO comment

sys.path.append("/home/nmiolane/code/")

sys.path.append("/home/nmiolane/code/ece594n/")

sys.path.append("/home/nmiolane/code/ece594n/hw_geomviz")

ninamiolane · 2023-02-13T16:48:47Z

tests/tests_geomstats/test_visualization_grassmannian.py

+matplotlib.use("Agg")  # NOQA
+
+
+class TestVisualizationGrassmanian(tests.conftest.TestCase):


the functions:

projection_to_two_d

projection_to_three_d
should be tested as well

and: two_d_to_projection

ninamiolane · 2023-02-13T16:49:00Z

geomstats/visualization/grassmannian.py

+from geomstats.geometry.symmetric_matrices import SymmetricMatrices
+
+
+def projection_to_two_d(projections, points):


Make this a method of the Grassmanian class below + test it in the test file

ninamiolane · 2023-02-13T16:49:18Z

geomstats/visualization/grassmannian.py

+    return np.asarray(storage)
+
+
+def projection_to_three_d(projections, points):


Make this a method of the Grassmanian class below + test it in the test file

ninamiolane · 2023-02-13T16:49:32Z

geomstats/visualization/grassmannian.py

+    return np.asarray(storage)
+
+
+def two_d_to_projection(vector):


Make this a method of the Grassmanian class below + test it.

ninamiolane · 2023-02-13T16:50:33Z

geomstats/visualization/grassmannian.py

+        Point.
+    point_b : array-like, shape=[..., dim]
+        Point.
+    metric : SpecialEuclideanMatrixCannonicalLeftMetric


Just say it is a RiemannianMetric

added grassmannian

1637a64

updated for linting

e224bc0

Christine8388811 added 11 commits January 28, 2023 13:30

fixed the import error in notebook

fdc1341

fixed an issue in the grassmannnian notebook

de66edc

fixed the syntax error in notebook

80ca12a

fixed linting

fe0ff23

fixed linting

9280a0e

fixed linting

02430f5

fixed linting

d367287

fixed linting

77853e2

fixed linting

e1028b3

changed test file name

fb8bfe6

Merge branch 'master' of https://github.com/geomstats/geomstats into …

d7596ce

…ChristineWan

christine051601 changed the title ~~added grassmannian~~ grassmannian test visuaiization for 2D Jan 30, 2023

ninamiolane added the ece594n label Feb 1, 2023

ninamiolane requested changes Feb 13, 2023

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

grassmannian test visuaiization for 2D #1803

grassmannian test visuaiization for 2D #1803

christine051601 commented Jan 28, 2023 •

edited

review-notebook-app bot commented Jan 28, 2023

codecov bot commented Jan 28, 2023 •

edited

ninamiolane left a comment

ninamiolane Feb 13, 2023

ninamiolane Feb 13, 2023

ninamiolane Feb 13, 2023

ninamiolane Feb 13, 2023

ninamiolane Feb 13, 2023

ninamiolane Feb 13, 2023

ninamiolane Feb 13, 2023

ninamiolane Feb 13, 2023

ninamiolane Feb 13, 2023

ninamiolane Feb 13, 2023

review-notebook-app bot commented Feb 13, 2023 •

edited

review-notebook-app bot commented Feb 13, 2023 •

edited

review-notebook-app bot commented Feb 13, 2023 •

edited

review-notebook-app bot commented Feb 13, 2023 •

edited

review-notebook-app bot commented Feb 13, 2023 •

edited

review-notebook-app bot commented Feb 13, 2023 •

edited

review-notebook-app bot commented Feb 13, 2023 •

edited

ninamiolane Feb 13, 2023

ninamiolane Feb 13, 2023

ninamiolane Feb 13, 2023

ninamiolane Feb 13, 2023

ninamiolane Feb 13, 2023

ninamiolane Feb 13, 2023

		return gs.concatenate([row_1, row_2], axis=-2)


		def _squared_d_g_a(point_a, point_b, metric):

		plt.plot(x_part, y_part, "-g")


		class GrassmannianCanonicalMetric(MatricesMetric, RiemannianMetric):

		matplotlib.use("Agg") # NOQA


		class TestVisualizationGrassmanian(tests.conftest.TestCase):

		from geomstats.geometry.symmetric_matrices import SymmetricMatrices


		def projection_to_two_d(projections, points):

		return np.asarray(storage)


		def projection_to_three_d(projections, points):

grassmannian test visuaiization for 2D #1803

Are you sure you want to change the base?

grassmannian test visuaiization for 2D #1803

Conversation

christine051601 commented Jan 28, 2023 • edited

review-notebook-app bot commented Jan 28, 2023

codecov bot commented Jan 28, 2023 • edited

Codecov Report

ninamiolane left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

review-notebook-app bot commented Feb 13, 2023 • edited

review-notebook-app bot commented Feb 13, 2023 • edited

review-notebook-app bot commented Feb 13, 2023 • edited

review-notebook-app bot commented Feb 13, 2023 • edited

review-notebook-app bot commented Feb 13, 2023 • edited

review-notebook-app bot commented Feb 13, 2023 • edited

review-notebook-app bot commented Feb 13, 2023 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

christine051601 commented Jan 28, 2023 •

edited

codecov bot commented Jan 28, 2023 •

edited

review-notebook-app bot commented Feb 13, 2023 •

edited

review-notebook-app bot commented Feb 13, 2023 •

edited

review-notebook-app bot commented Feb 13, 2023 •

edited

review-notebook-app bot commented Feb 13, 2023 •

edited

review-notebook-app bot commented Feb 13, 2023 •

edited

review-notebook-app bot commented Feb 13, 2023 •

edited

review-notebook-app bot commented Feb 13, 2023 •

edited