Binary serialization
NPY format
A simple format for saving numpy arrays to disk with the full information about them.
The .npy format is the standard binary file format in NumPy for persisting a single arbitrary NumPy array on disk. The format stores all of the shape and dtype information necessary to reconstruct the array correctly even on another machine with a different architecture. The format is designed to be as simple as possible while achieving its limited goals.
The .npz format is the standard format for persisting multiple NumPy arrays on disk. A .npz file is a zip file containing multiple .npy files, one for each array.
Capabilities
- Can represent all NumPy arrays including nested record arrays and object arrays.
- Represents the data in its native binary form.
- Supports Fortran-contiguous arrays directly.
- Stores all of the necessary information to reconstruct the array including shape and dtype on a machine of a different architecture. Both little-endian and big-endian arrays are supported, and a file with little-endian numbers will yield a little-endian array on any machine reading the file. The types are described in terms of their actual sizes. For example, if a machine with a 64-bit C "long int" writes out an array with "long ints", a reading machine with 32-bit C "long ints" will yield an array with 64-bit integers.
- Is straightforward to reverse engineer. Datasets often live longer than the programs that created them. A competent developer should be able to create a solution in their preferred programming language to read most .npy files that they have been given without much documentation.
- Allows memory-mapping of the data. See
open_memmap
. - Can be read from a filelike stream object instead of an actual file.
- Stores object arrays, i.e. arrays containing elements that are arbitrary Python objects. Files with object arrays are not to be mmapable, but can be read and written to disk.
Limitations
- Arbitrary subclasses of numpy.ndarray are not completely preserved. Subclasses will be accepted for writing, but only the array data will be written out. A regular numpy.ndarray object will be created upon reading the file.
Warning
Due to limitations in the interpretation of structured dtypes, dtypes with fields with empty names will have the names replaced by 'f0', 'f1', etc. Such arrays will not round-trip through the format entirely accurately. The data is intact; only the field names will differ. We are working on a fix for this. This fix will not require a change in the file format. The arrays with such structures can still be saved and restored, and the correct dtype may be restored by using the loadedarray.view(correct_dtype) method.
File extensions
We recommend using the .npy and .npz extensions for files saved in this format. This is by no means a requirement; applications may wish to use these file formats but use an extension specific to the application. In the absence of an obvious alternative, however, we suggest using .npy and .npz.
Version numbering
The version numbering of these formats is independent of NumPy version
numbering. If the format is upgraded, the code in numpy.io
will still
be able to read and write Version 1.0 files.
Format Version 1.0
The first 6 bytes are a magic string: exactly \x93NUMPY.
The next 1 byte is an unsigned byte: the major version number of the file format, e.g. \x01.
The next 1 byte is an unsigned byte: the minor version number of the file format, e.g. \x00. Note: the version of the file format is not tied to the version of the numpy package.
The next 2 bytes form a little-endian unsigned short int: the length of the header data HEADER_LEN.
The next HEADER_LEN bytes form the header data describing the array's format. It is an ASCII string which contains a Python literal expression of a dictionary. It is terminated by a newline (\n) and padded with spaces (\x20) to make the total of len(magic string) + 2 + len(length) + HEADER_LEN be evenly divisible by 64 for alignment purposes.
The dictionary contains three keys:
- "descr" : dtype.descr
- An object that can be passed as an argument to the
numpy.dtype
constructor to create the array's dtype.- "fortran_order" : bool
- Whether the array data is Fortran-contiguous or not. Since Fortran-contiguous arrays are a common form of non-C-contiguity, we allow them to be written directly to disk for efficiency.
- "shape" : tuple of int
- The shape of the array.
For repeatability and readability, the dictionary keys are sorted in alphabetic order. This is for convenience only. A writer SHOULD implement this if possible. A reader MUST NOT depend on this.
Following the header comes the array data. If the dtype contains Python objects (i.e. dtype.hasobject is True), then the data is a Python pickle of the array. Otherwise the data is the contiguous (either C- or Fortran-, depending on fortran_order) bytes of the array. Consumers can figure out the number of bytes by multiplying the number of elements given by the shape (noting that shape=() means there is 1 element) by dtype.itemsize.
Format Version 2.0
The version 1.0 format only allowed the array header to have a total size of
65535 bytes. This can be exceeded by structured arrays with a large number of
columns. The version 2.0 format extends the header size to 4 GiB.
numpy.save
will automatically save in 2.0 format if the data requires it,
else it will always use the more compatible 1.0 format.
The description of the fourth element of the header therefore has become: "The next 4 bytes form a little-endian unsigned int: the length of the header data HEADER_LEN."
Format Version 3.0
This version replaces the ASCII string (which in practice was latin1) with a utf8-encoded string, so supports structured types with any unicode field names.
Notes
The .npy format, including motivation for creating it and a comparison of alternatives, is described in the :doc:`"npy-format" NEP <neps:nep-0001-npy-format>`, however details have evolved with time and this document is more current.
Function | descr |
Returns a dtype based off the given description. |
Function | dtype |
Get a serializable descriptor from the dtype. |
Function | header |
Get the dictionary of header metadata from a numpy.ndarray. |
Function | magic |
Return the magic string for the given file format version. |
Function | open |
Open a .npy file as a memory-mapped array. |
Function | read |
Read an array from an NPY file. |
Function | read |
Read an array header from a filelike object using the 1.0 file format version. |
Function | read |
Read an array header from a filelike object using the 2.0 file format version. |
Function | read |
Read the magic string to get the version of the file format. |
Function | write |
Write an array to an NPY file, including a header. |
Function | write |
Write the header for an array using the 1.0 format. |
Function | write |
The 2.0 format allows storing very large structured arrays. |
Constant | ARRAY |
Undocumented |
Constant | BUFFER |
Undocumented |
Constant | EXPECTED |
Undocumented |
Constant | GROWTH |
Undocumented |
Constant | MAGIC |
Undocumented |
Constant | MAGIC |
Undocumented |
Function | _check |
Undocumented |
Function | _filter |
Clean up 'L' in npz header ints. |
Function | _has |
Undocumented |
Function | _read |
see read_array_header_1_0 |
Function | _read |
Read from file-like object until size bytes are read. Raises ValueError if not EOF is encountered before size bytes are read. Non-blocking objects only supported if they derive from io objects. |
Function | _wrap |
Takes a stringified header, and attaches the prefix and padding to it |
Function | _wrap |
Like _wrap_header , but chooses an appropriate version given the contents |
Function | _write |
Write the header for an array and returns the version used |
Constant | _MAX |
Undocumented |
Variable | _header |
Undocumented |
Returns a dtype based off the given description.
This is essentially the reverse of dtype_to_descr()
. It will remove
the valueless padding fields created by, i.e. simple fields like
dtype('float32'), and then convert the description to its corresponding
dtype.
Parameters | |
descr:object | The object retrieved by dtype.descr. Can be passed to
numpy.dtype() in order to replicate the input dtype. |
Returns | |
dtype | dtype - The dtype constructed by the description. |
Get a serializable descriptor from the dtype.
The .descr attribute of a dtype object cannot be round-tripped through the dtype() constructor. Simple types, like dtype('float32'), have a descr which looks like a record array with one field with '' as a name. The dtype() constructor interprets this as a request to give a default name. Instead, we construct descriptor that can be passed to dtype().
Parameters | |
dtype:dtype | The dtype of the array that will be written to disk. |
Returns | |
object | descr - An object that can be passed to numpy.dtype() in order to
replicate the input dtype. |
Get the dictionary of header metadata from a numpy.ndarray.
Parameters | |
array:numpy.ndarray | |
Returns | |
dict | d - This has the appropriate entries for writing its string representation to the header of the file. |
Return the magic string for the given file format version.
Parameters | |
major:int in[0, 255] | |
minor:int in[0, 255] | |
Returns | |
str | magic |
Raises | |
ValueError if the version cannot be formatted. |
Open a .npy file as a memory-mapped array.
This may be used to read an existing file or create a new one.
See Also
numpy.memmap
Parameters | |
filename:str or path-like | The name of the file on disk. This may not be a file-like object. |
mode:str , optional | The mode in which to open the file; the default is 'r+'. In
addition to the standard file modes, 'c' is also accepted to mean
"copy on write." See memmap for the available mode strings. |
dtype:data-type , optional | The data type of the array if we are creating a new file in "write"
mode, if not, dtype is ignored. The default value is None, which
results in a data-type of float64 . |
shape:tuple of int | The shape of the array if we are creating a new file in "write" mode, in which case this parameter is required. Otherwise, this parameter is ignored and is thus optional. |
fortranbool , optional | Whether the array should be Fortran-contiguous (True) or C-contiguous (False, the default) if we are creating a new file in "write" mode. |
version:tuple of int(major , minor) or None | If the mode is a "write" mode, then this is the version of the file format used to create the file. None means use the oldest supported version that is able to store the data. Default: None |
maxint , optional | Maximum allowed size of the header. Large headers may not be safe
to load securely and thus require explicitly passing a larger value.
See ast.literal_eval() for details. |
Returns | |
memmap | marray - The memory-mapped array. |
Raises | |
ValueError | If the data or the mode is invalid. |
OSError | If the file is not found or cannot be opened correctly. |
Read an array from an NPY file.
Parameters | |
fp:file_like object | If this is not a real file object, then this may take extra memory and time. |
allowbool , optional | Whether to allow writing pickled data. Default: False
Changed in version 1.16.3: Made default False in response to CVE-2019-6446.
|
pickledict | Additional keyword arguments to pass to pickle.load. These are only useful when loading object arrays saved on Python 2 when using Python 3. |
maxint , optional | Maximum allowed size of the header. Large headers may not be safe
to load securely and thus require explicitly passing a larger value.
See ast.literal_eval() for details.
This option is ignored when allow_pickle is passed. In that case
the file is by definition trusted and the limit is unnecessary. |
Returns | |
ndarray | array - The array from the data on disk. |
Raises | |
ValueError | If the data is invalid, or allow_pickle=False and the file contains an object array. |
Read an array header from a filelike object using the 1.0 file format version.
This will leave the file object located just after the header.
Parameters | |
fp:filelike object | A file object or something with a .read() method like a file. |
max | Undocumented |
Returns | |
| |
Raises | |
ValueError | If the data is invalid. |
Read an array header from a filelike object using the 2.0 file format version.
This will leave the file object located just after the header.
Parameters | |
fp:filelike object | A file object or something with a .read() method like a file. |
maxint , optional | Maximum allowed size of the header. Large headers may not be safe
to load securely and thus require explicitly passing a larger value.
See ast.literal_eval() for details. |
Returns | |
Raises | |
ValueError | If the data is invalid. |
Write an array to an NPY file, including a header.
If the array is neither C-contiguous nor Fortran-contiguous AND the file_like object is not a real file object, this function will have to copy data in memory.
Parameters | |
fp:file_like object | An open, writable file object, or similar object with a .write() method. |
array:ndarray | The array to write to disk. |
version:(int , int) or None , optional | The version number of the format. None means use the oldest supported version that is able to store the data. Default: None |
allowbool , optional | Whether to allow writing pickled data. Default: True |
pickledict , optional | Additional keyword arguments to pass to pickle.dump, excluding 'protocol'. These are only useful when pickling objects in object arrays on Python 3 to Python 2 compatible format. |
Raises | |
ValueError | If the array cannot be persisted. This includes the case of allow_pickle=False and array being an object array. |
Various other errors | If the array contains Python objects as part of its dtype, the process of pickling them may raise various errors if the objects are not picklable. |
Write the header for an array using the 1.0 format.
Parameters | |
fp:filelike object | |
d:dict | This has the appropriate entries for writing its string representation to the header of the file. |
- Write the header for an array using the 2.0 format.
- The 2.0 format allows storing very large structured arrays.
Parameters | |
fp:filelike object | |
d:dict | This has the appropriate entries for writing its string representation to the header of the file. |
Read from file-like object until size bytes are read. Raises ValueError if not EOF is encountered before size bytes are read. Non-blocking objects only supported if they derive from io objects.
Required as e.g. ZipExtFile in python 2.6 can return less data than requested.
Write the header for an array and returns the version used
Parameters | |
fp:filelike object | |
d:dict | This has the appropriate entries for writing its string representation to the header of the file. |
version:tuple or None | None means use oldest that works. Providing an explicit version will raise a ValueError if the format does not allow saving this data. Default: None |