Describe the bug
Enrichment factor function CalcEnrichment returns empty list, when it should return something, e.g. raise error, or return 0.

To Reproduce

Note: of course those values and fraction 0.9 make no particular sense. This was discovered when writing tests for edge cases.

import numpy as np
from rdkit.ML.Scoring.Scoring import CalcEnrichment

y_score = np.array([0.0, 0.0, 0.0, 0.0, 1.0])
y_test = np.array([0, 0, 0, 1, 1])
fraction = 0.9

scores = sorted(zip(y_score, y_test), reverse=True)
e = CalcEnrichment(scores, col=1, fractions=[fraction])
print(e)

The code returns an empty list, i.e. e = [].

I suspect this part here

  for i in range(numMol):
    if i > (numPerFrac[0] - 1) and i > 0:
      enrich.append(1.0 * numActives * numMol / i)
      numPerFrac.pop(0)
    active = scores[i][col]
    if active:
      numActives += 1

  if numActives > 0:  # check that there are actives
    enrich = [e / numActives for e in enrich]

For code above, numActives is 2, but enrich list is empty. This results in empty list comprehension, and returning an empty list.

Expected behavior
Maybe an error? Or zero value. However, even for those weird values there should be some enrichment, since a single active was found.

Configuration (please complete the following information):

• RDKit version: 2024.3.3
• OS: Ubuntu 24.04
• Python version (if relevant): 3.9.20
• Are you using conda? no
• If you are not using conda: how did you install the RDKit? pip