Methods
column_type(C) → {String}
Determine if the column is numeric, categorical, or other.
Parameters:
| Name | Type | Description |
|---|---|---|
C |
Array | 1d array of values, represents single column. |
- Source:
Returns:
Type of the output. Could be either 'category'
or 'number'.
- Type
- String
fit(X, y, feature_names)
Determines the type of columns, and calculates all the necessary
parameters for imputation of missing values and conversion of
categorical values, if any.
Parameters:
| Name | Type | Default | Description |
|---|---|---|---|
X |
Array | Matrix of raw inputs, where columns correspond to different features. | |
y |
Array | A vector of outputs. A type of problem could be determined from the type of the output. | |
feature_names |
Array | null | An array that contains names of the features as strings. |
- Source:
get_column(X, i)
Select a single column of matrix X.
Parameters:
| Name | Type | Description |
|---|---|---|
X |
Array | nested array, represents matrix of raw inputs. |
i |
Integer | Index of the column to select |
- Source:
transform(X, y)
Convert some inputs to purely numerical features, suitable for
further use in ML pipelines.
Parameters:
| Name | Type | Default | Description |
|---|---|---|---|
X |
Array | Matrix of raw inputs. These should be of the same format as the inputs used to `fit` method. | |
y |
Array | null | Vector of raw inputs. |
- Source: