Added MMD

[Negiiiin]

PR Type

[Feature ]

Short Description

Added MMD

---

This change is 

[afkanpour]

Reviewed all commit messages.
Reviewable status: 0 of 1 files reviewed, 7 unresolved discussions (waiting on @Negiiiin)

---

openpmcvl/statistical_analysis/MMD.py line 10 at r1 (raw file):

    From https://www.onurtunali.com/ml/2019/03/08/maximum-mean-discrepancy-in-machine-learning.html#formal-definition
"""
def MMD(x, y, kernel):

I believe the guideline says to use lowercase letters for function names.

---

openpmcvl/statistical_analysis/MMD.py line 15 at r1 (raw file):

    Args:
        x: first sample, distribution P

Please specify dimension of x and y in the docstring.

---

openpmcvl/statistical_analysis/MMD.py line 19 at r1 (raw file):

        kernel: kernel type such as "multiscale" or "rbf"
    """
    xx, yy, zz = torch.mm(x, x.t()), torch.mm(y, y.t()), torch.mm(x, y.t())

How about xy instead of zz since it is x y^T?

Code quote:

zz

---

openpmcvl/statistical_analysis/MMD.py line 23 at r1 (raw file):

    ry = (yy.diag().unsqueeze(0).expand_as(yy))

    dxx = rx.t() + rx - 2. * xx # Used for A in (1)

What is "(1)"?

---

openpmcvl/statistical_analysis/MMD.py line 23 at r1 (raw file):

    ry = (yy.diag().unsqueeze(0).expand_as(yy))

    dxx = rx.t() + rx - 2. * xx # Used for A in (1)

Please update these comments and indicate what intermediate quantities are calculated here.

Code quote:

# Used for A in (1)

---

openpmcvl/statistical_analysis/MMD.py line 27 at r1 (raw file):

    dxy = rx.t() + ry - 2. * zz # Used for C in (1)

    XX, YY, XY = (torch.zeros(xx.shape).to(device),

Please use lowercase variable names everywhere.

---

openpmcvl/statistical_analysis/MMD.py line 54 at r1 (raw file):

    """
    Loads all .pt files in a directory and combines them into a matrix.
    

Please remove all instances of trailing whitespaces.