Apply a smart bucketizer transformer
Apply a smart bucketizer transformer
label feature
option to keep track of values that were missing
option to keep track of invalid values, eg. NaN, -/+Inf or values that fall outside the buckets
minimum info gain, one of the stopping criteria of the Decision Tree
Apply NumericBucketizer transformer shortcut function
Apply NumericBucketizer transformer shortcut function
option to keep track of values that were missing
option to keep track of invalid values, eg. NaN, -/+Inf or values that fall outside the buckets
sorted list of split points for bucketizing
should the splits be left or right inclusive. Meaning if x1 and x2 are split points, then for Left the bucket interval is [x1, x2) and for Right the bucket interval is (x1, x2].
sorted list of labels for the buckets
FeatureLike
Fill missing values with mean
Fill missing values with mean
default value is the whole feature is filled with missing values
transformed feature of type RealNN
Apply real vectorizer: Converts a sequence of Real features into a vector feature.
Apply real vectorizer: Converts a sequence of Real features into a vector feature.
value to pull in place of nulls
replace missing values with mean (as apposed to constant provided in fillValue)
keep tract of when nulls occur by adding a second column to the vector with a null indicator
other features of same type
option to keep track of invalid values, eg. NaN, -/+Inf or values that fall outside the buckets
minimum info gain, one of the stopping criteria of the Decision Tree for the autoBucketizer
optional label column to be passed into autoBucketizer if present
a vector feature containing the raw Features with filled missing values and the bucketized features if a label argument is passed
Enrichment functions for Real Feature