mlrl.testbed.output_writer module¶
Author Michael Rapp (michael.rapp.ml@gmail.com)
Provides utilities for writing output data to sinks like the console or output files.
- class mlrl.testbed.output_writer.Formattable¶
Bases:
ABC
An abstract base class for all classes from which a textual representation can be created.
- class mlrl.testbed.output_writer.OutputWriter(sinks: List[Sink])¶
Bases:
ABC
An abstract base class for all classes that allow to write output data to one or several sinks, e.g., the console or output files.
- class CsvSink(output_dir: str, file_name: str, options: ~mlrl.common.options.Options = <mlrl.common.options.Options object>)¶
Bases:
Sink
Allows to write output data to a CSV file.
- write_output(meta_data: MetaData, data_split: DataSplit, data_type: DataType | None, prediction_scope: PredictionScope | None, output_data, **kwargs)¶
Must be implemented by subclasses in order to write output data to the sink.
- Parameters:
meta_data – The meta data of the data set
data_split – Information about the split of the available data, the output data corresponds to
data_type – Specifies whether the predictions and ground truth labels correspond to the training or test data or None, if no predictions have been obtained
prediction_scope – Specifies whether the predictions have been obtained from a global model or incrementally or None, if no predictions have been obtained
output_data – The output data that should be written to the sink
- KWARG_DATA_SPLIT = 'data_split'¶
- class LogSink(title: str, options: ~mlrl.common.options.Options = <mlrl.common.options.Options object>)¶
Bases:
Sink
Allows to write output data to the console.
- write_output(meta_data: MetaData, data_split: DataSplit, data_type: DataType | None, prediction_scope: PredictionScope | None, output_data, **kwargs)¶
Must be implemented by subclasses in order to write output data to the sink.
- Parameters:
meta_data – The meta data of the data set
data_split – Information about the split of the available data, the output data corresponds to
data_type – Specifies whether the predictions and ground truth labels correspond to the training or test data or None, if no predictions have been obtained
prediction_scope – Specifies whether the predictions have been obtained from a global model or incrementally or None, if no predictions have been obtained
output_data – The output data that should be written to the sink
- class Sink(options: ~mlrl.common.options.Options = <mlrl.common.options.Options object>)¶
Bases:
ABC
An abstract base class for all sinks, output data may be written to.
- abstract write_output(meta_data: MetaData, data_split: DataSplit, data_type: DataType | None, prediction_scope: PredictionScope | None, output_data, **kwargs)¶
Must be implemented by subclasses in order to write output data to the sink.
- Parameters:
meta_data – The meta data of the data set
data_split – Information about the split of the available data, the output data corresponds to
data_type – Specifies whether the predictions and ground truth labels correspond to the training or test data or None, if no predictions have been obtained
prediction_scope – Specifies whether the predictions have been obtained from a global model or incrementally or None, if no predictions have been obtained
output_data – The output data that should be written to the sink
- class TxtSink(output_dir: str, file_name: str, options: ~mlrl.common.options.Options = <mlrl.common.options.Options object>)¶
Bases:
Sink
Allows to write output data to a text file.
- write_output(meta_data: MetaData, data_split: DataSplit, data_type: DataType | None, prediction_scope: PredictionScope | None, output_data, **kwargs)¶
Must be implemented by subclasses in order to write output data to the sink.
- Parameters:
meta_data – The meta data of the data set
data_split – Information about the split of the available data, the output data corresponds to
data_type – Specifies whether the predictions and ground truth labels correspond to the training or test data or None, if no predictions have been obtained
prediction_scope – Specifies whether the predictions have been obtained from a global model or incrementally or None, if no predictions have been obtained
output_data – The output data that should be written to the sink
- write_output(meta_data: MetaData, x, y, data_split: DataSplit, learner, data_type: DataType | None = None, prediction_type: PredictionType | None = None, prediction_scope: PredictionScope | None = None, predictions: Any | None = None, train_time: float = 0, predict_time: float = 0)¶
Generates the output data and writes it to all available sinks.
- Parameters:
meta_data – The meta-data of the data set
x – A numpy.ndarray or scipy.sparse matrix, shape (num_examples, num_features), that stores the feature values
y – A numpy.ndarray or scipy.sparse matrix, shape (num_examples, num_labels), that stores the ground truth labels
data_split – Information about the split of the available data, the output data corresponds to
learner – The learner that has been trained
data_type – Specifies whether the predictions and ground truth labels correspond to the training or test data or None, if no predictions have been obtained
prediction_type – The type of the predictions or None, if no predictions have been obtained
prediction_scope – Specifies whether the predictions have been obtained from a global model or incrementally or None, if no predictions have been obtained
predictions – A numpy.ndarray or scipy.sparse matrix, shape (num_examples, num_labels), that stores the predictions for the query examples or None, if no predictions have been obtained
train_time – The time needed for training or 0, if no model has been trained
predict_time – The time needed for prediction or 0, if no predictions have been obtained