tech.ml.dataset.impl.column — techascent/tech.ml.dataset 2.15

->persistent-map clj

(->persistent-map item)

create-missing-reader clj macro

(create-missing-reader datatype missing data n-elems options)

create-object-missing-reader clj

(create-object-missing-reader missing data n-elems options)

create-string-text-missing-reader clj

(create-string-text-missing-reader missing data n-elems options)

datatype->missing-value clj

(datatype->missing-value dtype)

dtype->missing-val-map clj

ensure-column clj

(ensure-column item)

Convert an item to a column if at all possible. Currently columns either implement the required protocols

ensure-column-reader clj

(ensure-column-reader values-seq)

ensure-column-seq clj

(ensure-column-seq item-seq)

extend-column-with-empty clj

(extend-column-with-empty column n-empty)

generic-reader! clj macro

(generic-reader!)

make-container clj

(make-container dtype)

(make-container dtype n-elems)

new-column clj

(new-column name data)

(new-column name data metadata)

(new-column name data metadata missing)

Given a map of (something convertible to a long reader) missing indexes, (something convertible to a reader) data and a (string or keyword) name, return an implementation of enough of the column and datatype protocols to allow efficient columnwise operations of the rest of tech.ml.dataset

object-primitive-array-types clj

prepend-column-with-empty clj

(prepend-column-with-empty column n-empty)

process-object-primitive-array-data clj

(process-object-primitive-array-data obj-data)

scan-data-for-missing clj

(scan-data-for-missing coldata)

scan-object-data-for-missing clj

(scan-object-data-for-missing container-dtype obj-data)

scan-object-numeric-data-for-missing clj

(scan-object-numeric-data-for-missing container-dtype obj-data)