NumpyFile

Base class: Object

Represents a NumPy .npy or .npz file handle.

Use .npy files for a single direct array value and .npz files for named arrays.

Example

NumpyFile("values.npy", "w").write([1.0, 2.0, 3.0])
var values = NumpyFile("values.npy").read(as_real | as_list)
NumpyFile("archive.npz", "w").write("samples", [1, 2, 3])

Constructors

NumpyFile(const string path) -> NumpyFile

Prepares a NumPy file in read/write mode, creating it on first write if needed.

Parameters

• path: File path. Missing extensions default to .npy.

Example

var f = NumpyFile("values.npy")

NumpyFile(const string path, const string mode) -> NumpyFile

Prepares a NumPy file with an explicit mode.

Modes:

• "r": read-only, file must exist
• "r+": read/write, file must exist
• "w": read/write, create or truncate on first write
• "w-" or "x": create on first write, fail if it exists
• "a": read/write, create on first write if missing (default)

Parameters

• path: File path.
• mode: File open mode.

Example

var f = NumpyFile("archive.npz", "w")

Members

Name	Description
dtype	Returns the dtype descriptor of a named .npz array.
exists	Checks whether a named array exists in a .npz archive.
is_npy	Returns whether this handle targets a .npy file.
is_npz	Returns whether this handle targets a .npz file.
list_datasets	Lists named arrays in a .npz archive.
path	Returns the resolved file path.
rank	Returns the rank of a named .npz array.
read	Reads a named complex sparse matrix from a .npz archive.
shape	Returns the shape of a .npy array.
size	Returns the element count of a .npy array.
write	Writes a sparse matrix densely to a .npy file and returns the file for chaining.