Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[MRG] Disregard NaNs in preprocessing #177

Merged
merged 3 commits into from
Jan 9, 2020
Merged

[MRG] Disregard NaNs in preprocessing #177

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
27efa30
Disregard nans in preprocessing. Update docstrings with examples and …
Jan 7, 2020
4d80dc7
linter fix
Jan 7, 2020
340a635
update changelog
Jan 8, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
3 changes: 3 additions & 0 deletions CHANGELOG.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,6 +19,9 @@ local optima using a parameter `n_init` set to 1 by default
module have changed their names to `load_time_series_txt` and
`save_time_series_txt`. Old names can still be used but considered deprecated
and removed from the public API documentation for the sake of harmonization
* `TimeSeriesScalerMeanVariance` and `TimeSeriesScalerMinMax` now ignore any
NaNs when calling their respective `transform` methods in order to better
mirror scikit-learn's handling of missing data in preprocessing.

### Added

Expand Down
23 changes: 19 additions & 4 deletions tslearn/preprocessing.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -103,12 +103,20 @@ class TimeSeriesScalerMinMax(TransformerMixin):
-----
This method requires a dataset of equal-sized time series.

NaNs within a time series are ignored when calculating min and max.

Examples
--------
>>> TimeSeriesScalerMinMax(value_range=(1., 2.)).fit_transform([[0, 3, 6]])
array([[[1. ],
[1.5],
[2. ]]])
>>> TimeSeriesScalerMinMax(value_range=(1., 2.)).fit_transform(
... [[numpy.nan, 3, 6]]
... )
array([[[nan],
[1. ],
[2. ]]])
"""
def __init__(self, value_range=(0., 1.), min=None, max=None):
self.value_range = value_range
Expand Down Expand Up @@ -164,8 +172,8 @@ def transform(self, X, y=None, **kwargs):
" than maximum. Got %s." % str(self.value_range))

X_ = to_time_series_dataset(X)
min_t = numpy.min(X_, axis=1)[:, numpy.newaxis, :]
max_t = numpy.max(X_, axis=1)[:, numpy.newaxis, :]
min_t = numpy.nanmin(X_, axis=1)[:, numpy.newaxis, :]
max_t = numpy.nanmax(X_, axis=1)[:, numpy.newaxis, :]
range_t = max_t - min_t
nomin = (X_ - min_t) * (self.value_range[1] - self.value_range[0])
X_ = nomin / range_t + self.value_range[0]
Expand All @@ -188,13 +196,20 @@ class TimeSeriesScalerMeanVariance(TransformerMixin):
-----
This method requires a dataset of equal-sized time series.

NaNs within a time series are ignored when calculating mu and std.

Examples
--------
>>> TimeSeriesScalerMeanVariance(mu=0.,
... std=1.).fit_transform([[0, 3, 6]])
array([[[-1.22474487],
[ 0. ],
[ 1.22474487]]])
>>> TimeSeriesScalerMeanVariance(mu=0.,
... std=1.).fit_transform([[numpy.nan, 3, 6]])
array([[[nan],
[-1.],
[ 1.]]])
"""
def __init__(self, mu=0., std=1.):
self.mu_ = mu
Expand Down Expand Up @@ -231,8 +246,8 @@ def transform(self, X, **kwargs):
Rescaled time series dataset
"""
X_ = to_time_series_dataset(X)
mean_t = numpy.mean(X_, axis=1)[:, numpy.newaxis, :]
std_t = numpy.std(X_, axis=1)[:, numpy.newaxis, :]
mean_t = numpy.nanmean(X_, axis=1)[:, numpy.newaxis, :]
std_t = numpy.nanstd(X_, axis=1)[:, numpy.newaxis, :]
std_t[std_t == 0.] = 1.

X_ = (X_ - mean_t) * self.std_ / std_t + self.mu_
Expand Down