Complex synthetic dataset #25
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
ogrisel
reviewed
Dec 13, 2023
ogrisel
reviewed
Dec 14, 2023
examples/plot_complex_data.py
Outdated
Comment on lines
29
to
40
| DEFAULT_SHAPE_RANGES = ( | ||
| (0.7, 0.9), | ||
| (1.0, 1.0), | ||
| (2.0, 3.0), | ||
| ) | ||
|
|
||
| DEFAULT_SCALE_RANGES = ( | ||
| (1, 20), | ||
| (1, 10), | ||
| (1.5, 5), | ||
| ) | ||
| n_events = 3 |
There was a problem hiding this comment.
Let's use the default parameters of the synthetic solver in the example.
hazardous/data/_competing_weibull.py
Outdated
|
|
||
| frame = pd.concat([X, y], axis=1) | ||
| return Bunch(data=frame[X.columns], target=X[y.columns], frame=frame) | ||
| return Bunch(data=frame[X.columns], target=frame[y.columns], frame=frame) |
There was a problem hiding this comment.
Should we return y_censored instead of y (uncensored) in Bunch when return_X_y=False and return_uncensored_data=False (default)?
| degree_interaction=2, | ||
| random_state=0, | ||
| ): | ||
| rng = np.random.RandomState(random_state) |
There was a problem hiding this comment.
Suggested change
| rng = np.random.RandomState(random_state) | |
| rng = check_random_state(random_state) |
hazardous/data/_competing_weibull.py
Outdated
Comment on lines
245
to
255
| X, event_durations, duration_argmin = make_complex_features_with_sparse_matrix( | ||
| n_events, | ||
| n_samples, | ||
| base_scale, | ||
| shape_ranges, | ||
| scale_ranges, | ||
| n_features, | ||
| features_rate, | ||
| degree_interaction, | ||
| random_state, | ||
| ) |
There was a problem hiding this comment.
Suggested change
| X, event_durations, duration_argmin = make_complex_features_with_sparse_matrix( | |
| n_events, | |
| n_samples, | |
| base_scale, | |
| shape_ranges, | |
| scale_ranges, | |
| n_features, | |
| features_rate, | |
| degree_interaction, | |
| random_state, | |
| ) | |
| X, event_durations, duration_argmin = make_complex_features_with_sparse_matrix( | |
| n_events=n_events, | |
| n_samples=n_samples, | |
| base_scale=base_scale, | |
| shape_ranges=shape_ranges, | |
| scale_ranges=scale_ranges, | |
| n_features=n_features, | |
| features_rate=features_rate, | |
| degree_interaction=degree_interaction, | |
| random_state=random_state, | |
| ) |
hazardous/data/_competing_weibull.py
Outdated
Comment on lines
257
to
259
| X, event_durations, duration_argmin = make_simple_features( | ||
| n_events, n_samples, base_scale, shape_ranges, scale_ranges, random_state | ||
| ) |
There was a problem hiding this comment.
Suggested change
| X, event_durations, duration_argmin = make_simple_features( | |
| n_events, n_samples, base_scale, shape_ranges, scale_ranges, random_state | |
| ) | |
| X, event_durations, duration_argmin = make_simple_features( | |
| n_events=n_events, | |
| n_samples=n_samples, | |
| base_scale=base_scale, | |
| shape_ranges=shape_ranges, | |
| scale_ranges=scale_ranges, | |
| random_state=random_state, | |
| ) |
hazardous/data/_competing_weibull.py
Outdated
Comment on lines
186
to
187
| df_features = pd.DataFrame(rng.randn(n_samples, n_features)) | ||
| df_features.columns = [f"feature_{i}" for i in range(n_features)] |
There was a problem hiding this comment.
Nitpick
Suggested change
| df_features = pd.DataFrame(rng.randn(n_samples, n_features)) | |
| df_features.columns = [f"feature_{i}" for i in range(n_features)] | |
| columns = [f"feature_{i}" for i in range(n_features)] | |
| df_features = pd.DataFrame( | |
| rng.randn(n_samples, n_features), | |
| columns=columns, | |
| ) |
hazardous/data/_competing_weibull.py
Outdated
Comment on lines
114
to
115
| shape = df_shape_scale_star[f"shape_{event}"].copy() | ||
| scale = df_shape_scale_star[f"scale_{event}"].copy() |
There was a problem hiding this comment.
Why do we need to make copies?
hazardous/data/_competing_weibull.py
Outdated
| df_shape_scale_star[f"shape_{event}"] = ( | ||
| shape_min | ||
| + (shape_max - shape_min) | ||
| * expit(scaler.fit_transform(shape.values.reshape(-1, 1))) |
There was a problem hiding this comment.
We don't need numpy conversion here, we can pass directly df_shape_scale_star[[f"shape_{event}"]] to the transformer.
hazardous/data/_competing_weibull.py
Outdated
| censoring_relative_scale=1.5, | ||
| random_state=0, | ||
| ): | ||
| rng = np.random.RandomState(random_state) |
There was a problem hiding this comment.
Suggested change
| rng = np.random.RandomState(random_state) | |
| rng = check_random_state(random_state) |
hazardous/data/_competing_weibull.py
Outdated
| n_features=10, | ||
| features_rate=0.3, | ||
| degree_interaction=2, | ||
| independent=True, |
There was a problem hiding this comment.
To be more specific?
Suggested change
| independent=True, | |
| independent_censoring=True, |
hazardous/data/_competing_weibull.py
Outdated
| features_impact_censoring, | ||
| 0, | ||
| ) | ||
| df_censoring_params = censoring_relative_scale * X @ features_impact_censoring |
There was a problem hiding this comment.
The variable names are quite hard to understand at first glance. Maybe w_ something instead of features_impact_censoring and X_params instead of df_censoring_params?
adding 50% features marinal and 50% features interaction in the creation of w_star
…hanged, some tests were broken)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Creating a complex dataset.