TrnValTst
Namespace: Datasets
Functions and values
| Function or value | Description |
apply f trnValTst
Signature: f:(Dataset<'?175717> -> '?175718) -> trnValTst:TrnValTst<'?175717> -> '?175718 * '?175718 * '?175718
Type parameters: '?175717, '?175718 |
Applies the given function to the training, validation and test partions and return the results as a tuple (trn, val, tst). |
load filename
Signature: filename:string -> TrnValTst<'S>
Type parameters: 'S |
Loads a dataset from an HDF5 file. HDF5 folders called 'Trn', 'Val' and 'Tst' are used to store the dataset parations. |
ofDataset dataset
Signature: dataset:Dataset<'?175722> -> TrnValTst<'?175722>
Type parameters: '?175722 |
Creates the partitioning from the specified dataset using a split of 80%, 10% and 10% for the training, validation and test partitions respectively. |
ofDatasetWithRatios (...)
Signature: (trnRatio:float<MeasureOne> * valRatio:float<MeasureOne> * tstRatio:float<MeasureOne>) -> dataset:Dataset<'?175720> -> TrnValTst<'?175720>
Type parameters: '?175720 |
Creates the partitioning from the specified dataset using the specified training, validation, test splits. |
save filename this
Signature: filename:string -> this:TrnValTst<'S> -> unit
Type parameters: 'S |
Saves this dataset to disk in an HDF5 file. HDF5 folders called 'Trn', 'Val' and 'Tst' are used to store the dataset parations. |
toCuda this
Signature: this:TrnValTst<'S> -> TrnValTst<'S>
Type parameters: 'S |
Copies the given dataset to a CUDA GPU. |
toHost this
Signature: this:TrnValTst<'S> -> TrnValTst<'S>
Type parameters: 'S |
Copies the given dataset to the host. |
