tech.ml.dataset.categorical — techascent/tech.ml.dataset 1.0-alpha2

Dealing with categorical dataset data involves having two mapping systems. The first is a map of category to integer within the same column. The second is a 'one-hot' encoding where you generate more columns but those have a reduced number of possible categories, usually one categorical value per column.

build-categorical-map clj

(build-categorical-map dataset column-name-seq & [table-value-list])

Given a dataset and these columns, produce a label-map of column-name to specific categorical label-map.

build-one-hot-map clj

(build-one-hot-map dataset column-name-seq & [one-hot-table-args])

column-categorical-map clj

(column-categorical-map categorical-map new-dtype old-column)

Given a categorical map for a given column, produce a new column of the desired datatype with the values mapped to the table values.

column-one-hot-map clj

(column-one-hot-map one-hot-map new-dtype dataset column-name)

Using one hot map, produce Y new columns while removing existing column.

column-values->categorical clj

(column-values->categorical dataset src-column categorical-map)

Given a column encoded via either string->number or one-hot, reverse map to the a sequence of the original string column values.

inverse-map-categorical-col-fn clj

(inverse-map-categorical-col-fn src-column column-categorical-map)

inverse-map-categorical-columns clj

(inverse-map-categorical-columns dataset src-column column-categorical-map)

is-one-hot-label-map? clj

(is-one-hot-label-map? label-map)

make-string-table-from-table-args clj

(make-string-table-from-table-args table-value-list)

Make a mapping of value->index from a list of either string values or [valname idx] pairs. Returns map of value->index.