qml.data.Dataset

class Dataset(bind=None, *, data_name=None, identifiers=None, **attrs)

Bases: pennylane.data.base.mapper.MapperMixin, pennylane.data.base.dataset._DatasetTransform

Base class for Datasets.

Attributes

attr_info	Returns a mapping of the AttributeInfo for each of this dataset's attributes.
attrs	Returns all attributes of this Dataset.
bind	Return the HDF5 group that contains this dataset.
data_name	Returns the data name (category) of this dataset.
fields	A mapping of attribute names to their Attribute information.
identifiers	Returns this dataset's parameters.
info	Return metadata associated with this dataset.
type_id	Type identifier for this dataset.

attr_info

Returns a mapping of the AttributeInfo for each of this dataset’s attributes.

attrs

Returns all attributes of this Dataset.

bind

Return the HDF5 group that contains this dataset.

data_name

Returns the data name (category) of this dataset.

fields = mappingproxy({})

A mapping of attribute names to their Attribute information. Note that this contains attributes declared on the class, not attributes added to an instance. Use attrs to view all attributes on an instance.

identifiers

Returns this dataset’s parameters.

info

Return metadata associated with this dataset.

type_id = 'dataset'

Type identifier for this dataset. Used internally to load datasets from other datasets.

Methods

close()	Close the underlying dataset file.
list_attributes()	Returns a list of this dataset's attributes.
open(filepath[, mode])	Open existing dataset or create a new one at filepath.
read(source[, attributes, overwrite])	Load dataset from HDF5 file at filepath.
write(dest[, mode, attributes, overwrite])	Write dataset to HDF5 file at filepath.

close()

Close the underlying dataset file. The dataset will become inaccessible.

list_attributes()

Returns a list of this dataset’s attributes.

classmethod open(filepath, mode='r')

Open existing dataset or create a new one at filepath.

Parameters

• filepath – Path to dataset file

• mode – File handling mode. Possible values are “w-” (create, fail if file exists), “w” (create, overwrite existing), “a” (append existing, create if doesn’t exist), “r” (read existing, must exist), and “copy”, which loads the dataset into memory and detaches it from the underlying file. Default is “r”.

Returns

Dataset object from file

read(source, attributes=None, *, overwrite=False)

Load dataset from HDF5 file at filepath.

Parameters

• source – Dataset, or path to HDF5 file containing dataset, from which to read attributes

• attributes – Optional list of attributes to copy. If None, all attributes will be copied.

• overwrite – Whether to overwrite attributes that already exist in this dataset.

write(dest, mode='a', attributes=None, *, overwrite=False)

Write dataset to HDF5 file at filepath.

Parameters

• dest – HDF5 file, or path to HDF5 file containing dataset, to write attributes to

• mode – File handling mode, if source is a file system path. Possible values are “w-” (create, fail if file exists), “w” (create, overwrite existing), and “a” (append existing, create if doesn’t exist). Default is “w-“.

• attributes – Optional list of attributes to copy. If None, all attributes will be copied. Note that identifiers will always be copied.

• overwrite – Whether to overwrite attributes that already exist in this dataset.

code/api/pennylane.data.Dataset

 

Download Python script

 

Download Notebook

 

View on GitHub