lightautoml.transformers
Basic feature generation steps and helper utils.
Base Classes
Base class for transformer method (like sklearn, but works with datasets). |
|
Transformer that contains the list of transformers and apply one by one sequentially. |
|
Transformer that apply the sequence on transformers in parallel on dataset and concatenate the result. |
|
Select columns to pass to another transformers (or feature selection). |
|
Apply 1 columns transformer to all columns. |
|
Apply multiple transformers and select best. |
|
Convert dataset to given type. |
|
Change data roles (include dtypes etc). |
Numeric
Create NaN flags. |
|
Fillna with median. |
|
Fill inf with nan to handle as nan value. |
|
Convert probs to logodds. |
|
Classic StandardScaler. |
|
Discretization of numeric features by quantiles. |
Categorical
Simple LabelEncoder in order of frequency. |
|
Simple OneHotEncoder over label encoded categories. |
|
Labels are encoded with frequency in train data. |
|
Encoding ordinal categories into numbers. |
|
Out-of-fold target encoding. |
|
Out-of-fold target encoding for multiclass task. |
|
Build label encoded intertsections of categorical variables. |
Datetime
Basic conversion strategy, used in selection one-to-one transformers. |
|
Basic conversion strategy, used in selection one-to-one transformers. |
|
Basic conversion strategy, used in selection one-to-one transformers. |
Decompositions
PCA. |
|
TruncatedSVD. |
Text
Base class for ML transformers. |
|
Simple Tfidf vectorizer. |
|
Simple tokenizer transformer. |
|
Out-of-fold sgd model prediction to reduce dimension of encoded text data. |
|
Concat text features transformer. |
|
Calculate text embeddings. |
Image
Simple image histogram. |
|
Calculate image embeddings. |