torch_frame
The semantic type of a column.
A semantic type denotes the semantic meaning of a column, and denotes how columns are encoded into an embedding space within tabular deep learning models:
import torch_frame
stype = torch_frame.numerical # Numerical columns
stype = torch_frame.categorical # Categorical columns
...
- numerical
Numerical columns.
- categorical
Categorical columns.
- text_embedded
Pre-computed embeddings of text columns.
- text_tokenized
Tokenized text columns for finetuning.
- multicategorical
Multicategorical columns.
- sequence_numerical
Sequence of numerical values.
- embedding
Embedding columns.
- timestamp
Timestamp columns.
- image_embedded
Pre-computed embeddings of image columns.
- class TaskType(value)[source]
The type of the task.
- REGRESSION
Regression task.
- MULTICLASS_CLASSIFICATION
Multi-class classification task.
- BINARY_CLASSIFICATION
Binary classification task.
- class NAStrategy(value)[source]
Strategy for dealing with NaN values in columns.
- MEAN
Replaces NaN values with the mean of a
torch_frame.numerical
column.
- ZEROS
Replaces NaN values with zeros in a
torch_frame.numerical
column.
- MOST_FREQUENT
Replaces NaN values with the most frequent category of a
torch_frame.categorical
column.