WrightTools.data.Channel.trim
- Channel.trim(neighborhood, method='ztest', factor=3, replace='nan', verbose=True)[source]
Remove outliers from the dataset.
Identifies outliers by comparing each point to its neighbors using a statistical test.
- Parameters:
neighborhood (list of integers) – Size of the neighborhood in each dimension. Length of the list must be equal to the dimensionality of the channel.
method ({'ztest'} (optional)) –
Statistical test used to detect outliers. Default is ztest.
- ztest
Compare point deviation from neighborhood mean to neighborhood standard deviation.
factor (number (optional)) – Tolerance factor. Default is 3.
replace ({'nan', 'mean', 'exclusive_mean', number} (optional)) –
Behavior of outlier replacement. Default is nan.
- nan
Outliers are replaced by numpy nans.
- mean
Outliers are replaced by the mean of its neighborhood, including itself.
- exclusive_mean
Outilers are replaced by the mean of its neighborhood, not including itself.
- number
Array becomes given number.
- Returns:
Indicies of trimmed outliers.
- Return type:
list of tuples
See also
clip
Remove pixels outside of a certain range.