diff options
author | Gard Spreemann <gspr@nonempty.org> | 2021-02-28 18:40:46 +0100 |
---|---|---|
committer | Gard Spreemann <gspr@nonempty.org> | 2021-02-28 21:50:48 +0100 |
commit | fddeb5724fe2e7f1f37476c5e3cfade992a4edec (patch) | |
tree | 41f77294950c7a3cdb793c0403e3be9c84ae4976 /src/python/doc/persistent_cohomology_sum.inc | |
parent | ccb63b32bc65c0a6030dfab0b70ece62d9eff988 (diff) |
Behave in line with scikit-learn guidelines
According to [1], we should in particular not do any validation in the
constructor, and fit/fit_transform should always update underscored
attributes (self.grid_ in this case).
We still want to allow for a user-defined, data-independent grid, so
we make this a separate parameter predefined_grid.
[1] https://scikit-learn.org/stable/developers/develop.html
Diffstat (limited to 'src/python/doc/persistent_cohomology_sum.inc')
0 files changed, 0 insertions, 0 deletions