cljml-core-matrix.core — cljml-core-matrix 0.1.0

Convert core.matrix datasets to weka datasets and vice versa, load/save arff files and other utilities

get-column-descr clj

(get-column-descr dset)

(get-column-descr dset colset)

Returns a vector of column names, if a column has nominal values, its entry is a map where key is the column name and value is a vector of the nominal values.

Clj-ml datasets require the column type to be specified in advance. Columns with categorical values need to have all the values specified in a vector

load-corematrix-dataset clj

(load-corematrix-dataset file-format filepath)

Returns a core.matrix dataset read from file. The file formats supported are :arff, :csv, :xrff and :libsvm

save-corematrix-dataset clj

(save-corematrix-dataset file-format dset filepath)

(save-corematrix-dataset file-format dset filepath dataset-name)

Saves the core.matrix dataset. The file formats supported are :arff, :csv, :xrff and :libsvm

to-cljml-dataset clj

(to-cljml-dataset dset dset-name)

(to-cljml-dataset dset dset-name col-descr)

Returns a clj-ml dataset given a core.matrix dataset and dataset name Optional argument is the column description for each column

to-corematrix-dataset clj

(to-corematrix-dataset dset)

Convert a cljml dataset to a core.matrix dataset

type-coerce clj

(type-coerce dset)

(type-coerce dset string-replace-fn)

cljml datasets need all values in a column to have the same type. NA values can be represented as nil, not empty strings. This method returns the dataset and replaces strings with nils