Alice Le Brigant

I agree we need to rethink the way we handle different quotients. Maybe the best way, from a user-perspective, would be to have a class `SRVShapeBundle` with boolean attributes `quotient_param`,...

To test sectional curvature more precisely than just the sign, you can check that for beta distributions, at any point `gs.array([x, y])`, it is given by ``` K_xy = gs.polygamma(2,...

Hi everyone, Thanks Luis for this great start ! I think it's really important that we can control in a simple fashion the way exp, log and geodesics are computed...

Great work ! @Jules-Deschamps did you check the compatibility of the hard coded metric matrices in classes like `DirichletDistributions` with the method `metric_matrix` of `FisherRaoMetric` you implemented ? To do...

@ninamiolane In the case "Frechet Mean for 152th cell shape with itself", the failure probably comes from the fact that the curve has two consecutive points that are the same,...

Ok I was not aware of this guideline ! Then I guess we should instead add "compute" of other verbs in infinite form where they are missing.

I have fixed the bug in the `align` method of `ShapeSrvBundle` in PR #1823. However, the way that random points are generated on the space of discrete curves should also...

Working example: ``` import matplotlib.pyplot as plt import geomstats.backend as gs from geomstats.geometry.discrete_curves import R2, SRVShapeBundle parametrized_curve_1 = lambda x: gs.transpose( gs.array([1 + 2 * gs.sin(gs.pi * x), 3 +...