|
|
import sysimport osimport reimport functoolsimport itertoolsimport warningsimport weakrefimport contextlibfrom operator import itemgetter, index as opindexfrom collections.abc import Mapping
import numpy as npfrom . import formatfrom ._datasource import DataSourcefrom numpy.core import overridesfrom numpy.core.multiarray import packbits, unpackbitsfrom numpy.core.overrides import set_array_function_like_doc, set_modulefrom numpy.core._internal import recursivefrom ._iotools import ( LineSplitter, NameValidator, StringConverter, ConverterError, ConverterLockError, ConversionWarning, _is_string_like, has_nested_fields, flatten_dtype, easy_dtype, _decode_line )
from numpy.compat import ( asbytes, asstr, asunicode, os_fspath, os_PathLike, pickle )
@set_module('numpy')def loads(*args, **kwargs): # NumPy 1.15.0, 2017-12-10 warnings.warn( "np.loads is deprecated, use pickle.loads instead", DeprecationWarning, stacklevel=2) return pickle.loads(*args, **kwargs)
__all__ = [ 'savetxt', 'loadtxt', 'genfromtxt', 'ndfromtxt', 'mafromtxt', 'recfromtxt', 'recfromcsv', 'load', 'loads', 'save', 'savez', 'savez_compressed', 'packbits', 'unpackbits', 'fromregex', 'DataSource' ]
array_function_dispatch = functools.partial( overrides.array_function_dispatch, module='numpy')
class BagObj: """
BagObj(obj)
Convert attribute look-ups to getitems on the object passed in.
Parameters ---------- obj : class instance Object on which attribute look-up is performed.
Examples -------- >>> from numpy.lib.npyio import BagObj as BO >>> class BagDemo: ... def __getitem__(self, key): # An instance of BagObj(BagDemo) ... # will call this method when any ... # attribute look-up is required ... result = "Doesn't matter what you want, " ... return result + "you're gonna get this" ... >>> demo_obj = BagDemo() >>> bagobj = BO(demo_obj) >>> bagobj.hello_there "Doesn't matter what you want, you're gonna get this" >>> bagobj.I_can_be_anything "Doesn't matter what you want, you're gonna get this"
"""
def __init__(self, obj): # Use weakref to make NpzFile objects collectable by refcount self._obj = weakref.proxy(obj)
def __getattribute__(self, key): try: return object.__getattribute__(self, '_obj')[key] except KeyError: raise AttributeError(key) from None
def __dir__(self): """
Enables dir(bagobj) to list the files in an NpzFile.
This also enables tab-completion in an interpreter or IPython. """
return list(object.__getattribute__(self, '_obj').keys())
def zipfile_factory(file, *args, **kwargs): """
Create a ZipFile.
Allows for Zip64, and the `file` argument can accept file, str, or pathlib.Path objects. `args` and `kwargs` are passed to the zipfile.ZipFile constructor. """
if not hasattr(file, 'read'): file = os_fspath(file) import zipfile kwargs['allowZip64'] = True return zipfile.ZipFile(file, *args, **kwargs)
class NpzFile(Mapping): """
NpzFile(fid)
A dictionary-like object with lazy-loading of files in the zipped archive provided on construction.
`NpzFile` is used to load files in the NumPy ``.npz`` data archive format. It assumes that files in the archive have a ``.npy`` extension, other files are ignored.
The arrays and file strings are lazily loaded on either getitem access using ``obj['key']`` or attribute lookup using ``obj.f.key``. A list of all files (without ``.npy`` extensions) can be obtained with ``obj.files`` and the ZipFile object itself using ``obj.zip``.
Attributes ---------- files : list of str List of all files in the archive with a ``.npy`` extension. zip : ZipFile instance The ZipFile object initialized with the zipped archive. f : BagObj instance An object on which attribute can be performed as an alternative to getitem access on the `NpzFile` instance itself. allow_pickle : bool, optional Allow loading pickled data. Default: False
.. versionchanged:: 1.16.3 Made default False in response to CVE-2019-6446.
pickle_kwargs : dict, optional Additional keyword arguments to pass on to pickle.load. These are only useful when loading object arrays saved on Python 2 when using Python 3.
Parameters ---------- fid : file or str The zipped archive to open. This is either a file-like object or a string containing the path to the archive. own_fid : bool, optional Whether NpzFile should close the file handle. Requires that `fid` is a file-like object.
Examples -------- >>> from tempfile import TemporaryFile >>> outfile = TemporaryFile() >>> x = np.arange(10) >>> y = np.sin(x) >>> np.savez(outfile, x=x, y=y) >>> _ = outfile.seek(0)
>>> npz = np.load(outfile) >>> isinstance(npz, np.lib.io.NpzFile) True >>> sorted(npz.files) ['x', 'y'] >>> npz['x'] # getitem access array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9]) >>> npz.f.x # attribute lookup array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
"""
# Make __exit__ safe if zipfile_factory raises an exception zip = None fid = None
def __init__(self, fid, own_fid=False, allow_pickle=False, pickle_kwargs=None): # Import is postponed to here since zipfile depends on gzip, an # optional component of the so-called standard library. _zip = zipfile_factory(fid) self._files = _zip.namelist() self.files = [] self.allow_pickle = allow_pickle self.pickle_kwargs = pickle_kwargs for x in self._files: if x.endswith('.npy'): self.files.append(x[:-4]) else: self.files.append(x) self.zip = _zip self.f = BagObj(self) if own_fid: self.fid = fid
def __enter__(self): return self
def __exit__(self, exc_type, exc_value, traceback): self.close()
def close(self): """
Close the file.
"""
if self.zip is not None: self.zip.close() self.zip = None if self.fid is not None: self.fid.close() self.fid = None self.f = None # break reference cycle
def __del__(self): self.close()
# Implement the Mapping ABC def __iter__(self): return iter(self.files)
def __len__(self): return len(self.files)
def __getitem__(self, key): # FIXME: This seems like it will copy strings around # more than is strictly necessary. The zipfile # will read the string and then # the format.read_array will copy the string # to another place in memory. # It would be better if the zipfile could read # (or at least uncompress) the data # directly into the array memory. member = False if key in self._files: member = True elif key in self.files: member = True key += '.npy' if member: bytes = self.zip.open(key) magic = bytes.read(len(format.MAGIC_PREFIX)) bytes.close() if magic == format.MAGIC_PREFIX: bytes = self.zip.open(key) return format.read_array(bytes, allow_pickle=self.allow_pickle, pickle_kwargs=self.pickle_kwargs) else: return self.zip.read(key) else: raise KeyError("%s is not a file in the archive" % key)
# deprecate the python 2 dict apis that we supported by accident in # python 3. We forgot to implement itervalues() at all in earlier # versions of numpy, so no need to deprecated it here.
def iteritems(self): # Numpy 1.15, 2018-02-20 warnings.warn( "NpzFile.iteritems is deprecated in python 3, to match the " "removal of dict.itertems. Use .items() instead.", DeprecationWarning, stacklevel=2) return self.items()
def iterkeys(self): # Numpy 1.15, 2018-02-20 warnings.warn( "NpzFile.iterkeys is deprecated in python 3, to match the " "removal of dict.iterkeys. Use .keys() instead.", DeprecationWarning, stacklevel=2) return self.keys()
@set_module('numpy')def load(file, mmap_mode=None, allow_pickle=False, fix_imports=True, encoding='ASCII'): """
Load arrays or pickled objects from ``.npy``, ``.npz`` or pickled files.
.. warning:: Loading files that contain object arrays uses the ``pickle`` module, which is not secure against erroneous or maliciously constructed data. Consider passing ``allow_pickle=False`` to load data that is known not to contain object arrays for the safer handling of untrusted sources.
Parameters ---------- file : file-like object, string, or pathlib.Path The file to read. File-like objects must support the ``seek()`` and ``read()`` methods. Pickled files require that the file-like object support the ``readline()`` method as well. mmap_mode : {None, 'r+', 'r', 'w+', 'c'}, optional If not None, then memory-map the file, using the given mode (see `numpy.memmap` for a detailed description of the modes). A memory-mapped array is kept on disk. However, it can be accessed and sliced like any ndarray. Memory mapping is especially useful for accessing small fragments of large files without reading the entire file into memory. allow_pickle : bool, optional Allow loading pickled object arrays stored in npy files. Reasons for disallowing pickles include security, as loading pickled data can execute arbitrary code. If pickles are disallowed, loading object arrays will fail. Default: False
.. versionchanged:: 1.16.3 Made default False in response to CVE-2019-6446.
fix_imports : bool, optional Only useful when loading Python 2 generated pickled files on Python 3, which includes npy/npz files containing object arrays. If `fix_imports` is True, pickle will try to map the old Python 2 names to the new names used in Python 3. encoding : str, optional What encoding to use when reading Python 2 strings. Only useful when loading Python 2 generated pickled files in Python 3, which includes npy/npz files containing object arrays. Values other than 'latin1', 'ASCII', and 'bytes' are not allowed, as they can corrupt numerical data. Default: 'ASCII'
Returns ------- result : array, tuple, dict, etc. Data stored in the file. For ``.npz`` files, the returned instance of NpzFile class must be closed to avoid leaking file descriptors.
Raises ------ IOError If the input file does not exist or cannot be read. ValueError The file contains an object array, but allow_pickle=False given.
See Also -------- save, savez, savez_compressed, loadtxt memmap : Create a memory-map to an array stored in a file on disk. lib.format.open_memmap : Create or load a memory-mapped ``.npy`` file.
Notes ----- - If the file contains pickle data, then whatever object is stored in the pickle is returned. - If the file is a ``.npy`` file, then a single array is returned. - If the file is a ``.npz`` file, then a dictionary-like object is returned, containing ``{filename: array}`` key-value pairs, one for each file in the archive. - If the file is a ``.npz`` file, the returned value supports the context manager protocol in a similar fashion to the open function::
with load('foo.npz') as data: a = data['a']
The underlying file descriptor is closed when exiting the 'with' block.
Examples -------- Store data to disk, and load it again:
>>> np.save('/tmp/123', np.array([[1, 2, 3], [4, 5, 6]])) >>> np.load('/tmp/123.npy') array([[1, 2, 3], [4, 5, 6]])
Store compressed data to disk, and load it again:
>>> a=np.array([[1, 2, 3], [4, 5, 6]]) >>> b=np.array([1, 2]) >>> np.savez('/tmp/123.npz', a=a, b=b) >>> data = np.load('/tmp/123.npz') >>> data['a'] array([[1, 2, 3], [4, 5, 6]]) >>> data['b'] array([1, 2]) >>> data.close()
Mem-map the stored array, and then access the second row directly from disk:
>>> X = np.load('/tmp/123.npy', mmap_mode='r') >>> X[1, :] memmap([4, 5, 6])
"""
if encoding not in ('ASCII', 'latin1', 'bytes'): # The 'encoding' value for pickle also affects what encoding # the serialized binary data of NumPy arrays is loaded # in. Pickle does not pass on the encoding information to # NumPy. The unpickling code in numpy.core.multiarray is # written to assume that unicode data appearing where binary # should be is in 'latin1'. 'bytes' is also safe, as is 'ASCII'. # # Other encoding values can corrupt binary data, and we # purposefully disallow them. For the same reason, the errors= # argument is not exposed, as values other than 'strict' # result can similarly silently corrupt numerical data. raise ValueError("encoding must be 'ASCII', 'latin1', or 'bytes'")
pickle_kwargs = dict(encoding=encoding, fix_imports=fix_imports)
with contextlib.ExitStack() as stack: if hasattr(file, 'read'): fid = file own_fid = False else: fid = stack.enter_context(open(os_fspath(file), "rb")) own_fid = True
# Code to distinguish from NumPy binary files and pickles. _ZIP_PREFIX = b'PK\x03\x04' _ZIP_SUFFIX = b'PK\x05\x06' # empty zip files start with this N = len(format.MAGIC_PREFIX) magic = fid.read(N) # If the file size is less than N, we need to make sure not # to seek past the beginning of the file fid.seek(-min(N, len(magic)), 1) # back-up if magic.startswith(_ZIP_PREFIX) or magic.startswith(_ZIP_SUFFIX): # zip-file (assume .npz) # Potentially transfer file ownership to NpzFile stack.pop_all() ret = NpzFile(fid, own_fid=own_fid, allow_pickle=allow_pickle, pickle_kwargs=pickle_kwargs) return ret elif magic == format.MAGIC_PREFIX: # .npy file if mmap_mode: return format.open_memmap(file, mode=mmap_mode) else: return format.read_array(fid, allow_pickle=allow_pickle, pickle_kwargs=pickle_kwargs) else: # Try a pickle if not allow_pickle: raise ValueError("Cannot load file containing pickled data " "when allow_pickle=False") try: return pickle.load(fid, **pickle_kwargs) except Exception as e: raise IOError( "Failed to interpret file %s as a pickle" % repr(file)) from e
def _save_dispatcher(file, arr, allow_pickle=None, fix_imports=None): return (arr,)
@array_function_dispatch(_save_dispatcher)def save(file, arr, allow_pickle=True, fix_imports=True): """
Save an array to a binary file in NumPy ``.npy`` format.
Parameters ---------- file : file, str, or pathlib.Path File or filename to which the data is saved. If file is a file-object, then the filename is unchanged. If file is a string or Path, a ``.npy`` extension will be appended to the filename if it does not already have one. arr : array_like Array data to be saved. allow_pickle : bool, optional Allow saving object arrays using Python pickles. Reasons for disallowing pickles include security (loading pickled data can execute arbitrary code) and portability (pickled objects may not be loadable on different Python installations, for example if the stored objects require libraries that are not available, and not all pickled data is compatible between Python 2 and Python 3). Default: True fix_imports : bool, optional Only useful in forcing objects in object arrays on Python 3 to be pickled in a Python 2 compatible way. If `fix_imports` is True, pickle will try to map the new Python 3 names to the old module names used in Python 2, so that the pickle data stream is readable with Python 2.
See Also -------- savez : Save several arrays into a ``.npz`` archive savetxt, load
Notes ----- For a description of the ``.npy`` format, see :py:mod:`numpy.lib.format`.
Any data saved to the file is appended to the end of the file.
Examples -------- >>> from tempfile import TemporaryFile >>> outfile = TemporaryFile()
>>> x = np.arange(10) >>> np.save(outfile, x)
>>> _ = outfile.seek(0) # Only needed here to simulate closing & reopening file >>> np.load(outfile) array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
>>> with open('test.npy', 'wb') as f: ... np.save(f, np.array([1, 2])) ... np.save(f, np.array([1, 3])) >>> with open('test.npy', 'rb') as f: ... a = np.load(f) ... b = np.load(f) >>> print(a, b) # [1 2] [1 3] """
if hasattr(file, 'write'): file_ctx = contextlib.nullcontext(file) else: file = os_fspath(file) if not file.endswith('.npy'): file = file + '.npy' file_ctx = open(file, "wb")
with file_ctx as fid: arr = np.asanyarray(arr) format.write_array(fid, arr, allow_pickle=allow_pickle, pickle_kwargs=dict(fix_imports=fix_imports))
def _savez_dispatcher(file, *args, **kwds): yield from args yield from kwds.values()
@array_function_dispatch(_savez_dispatcher)def savez(file, *args, **kwds): """Save several arrays into a single file in uncompressed ``.npz`` format.
Provide arrays as keyword arguments to store them under the corresponding name in the output file: ``savez(fn, x=x, y=y)``.
If arrays are specified as positional arguments, i.e., ``savez(fn, x, y)``, their names will be `arr_0`, `arr_1`, etc.
Parameters ---------- file : str or file Either the filename (string) or an open file (file-like object) where the data will be saved. If file is a string or a Path, the ``.npz`` extension will be appended to the filename if it is not already there. args : Arguments, optional Arrays to save to the file. Please use keyword arguments (see `kwds` below) to assign names to arrays. Arrays specified as args will be named "arr_0", "arr_1", and so on. kwds : Keyword arguments, optional Arrays to save to the file. Each array will be saved to the output file with its corresponding keyword name.
Returns ------- None
See Also -------- save : Save a single array to a binary file in NumPy format. savetxt : Save an array to a file as plain text. savez_compressed : Save several arrays into a compressed ``.npz`` archive
Notes ----- The ``.npz`` file format is a zipped archive of files named after the variables they contain. The archive is not compressed and each file in the archive contains one variable in ``.npy`` format. For a description of the ``.npy`` format, see :py:mod:`numpy.lib.format`.
When opening the saved ``.npz`` file with `load` a `NpzFile` object is returned. This is a dictionary-like object which can be queried for its list of arrays (with the ``.files`` attribute), and for the arrays themselves.
When saving dictionaries, the dictionary keys become filenames inside the ZIP archive. Therefore, keys should be valid filenames. E.g., avoid keys that begin with ``/`` or contain ``.``.
Examples -------- >>> from tempfile import TemporaryFile >>> outfile = TemporaryFile() >>> x = np.arange(10) >>> y = np.sin(x)
Using `savez` with \\*args, the arrays are saved with default names.
>>> np.savez(outfile, x, y) >>> _ = outfile.seek(0) # Only needed here to simulate closing & reopening file >>> npzfile = np.load(outfile) >>> npzfile.files ['arr_0', 'arr_1'] >>> npzfile['arr_0'] array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
Using `savez` with \\**kwds, the arrays are saved with the keyword names.
>>> outfile = TemporaryFile() >>> np.savez(outfile, x=x, y=y) >>> _ = outfile.seek(0) >>> npzfile = np.load(outfile) >>> sorted(npzfile.files) ['x', 'y'] >>> npzfile['x'] array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
"""
_savez(file, args, kwds, False)
def _savez_compressed_dispatcher(file, *args, **kwds): yield from args yield from kwds.values()
@array_function_dispatch(_savez_compressed_dispatcher)def savez_compressed(file, *args, **kwds): """
Save several arrays into a single file in compressed ``.npz`` format.
Provide arrays as keyword arguments to store them under the corresponding name in the output file: ``savez(fn, x=x, y=y)``.
If arrays are specified as positional arguments, i.e., ``savez(fn, x, y)``, their names will be `arr_0`, `arr_1`, etc.
Parameters ---------- file : str or file Either the filename (string) or an open file (file-like object) where the data will be saved. If file is a string or a Path, the ``.npz`` extension will be appended to the filename if it is not already there. args : Arguments, optional Arrays to save to the file. Please use keyword arguments (see `kwds` below) to assign names to arrays. Arrays specified as args will be named "arr_0", "arr_1", and so on. kwds : Keyword arguments, optional Arrays to save to the file. Each array will be saved to the output file with its corresponding keyword name.
Returns ------- None
See Also -------- numpy.save : Save a single array to a binary file in NumPy format. numpy.savetxt : Save an array to a file as plain text. numpy.savez : Save several arrays into an uncompressed ``.npz`` file format numpy.load : Load the files created by savez_compressed.
Notes ----- The ``.npz`` file format is a zipped archive of files named after the variables they contain. The archive is compressed with ``zipfile.ZIP_DEFLATED`` and each file in the archive contains one variable in ``.npy`` format. For a description of the ``.npy`` format, see :py:mod:`numpy.lib.format`.
When opening the saved ``.npz`` file with `load` a `NpzFile` object is returned. This is a dictionary-like object which can be queried for its list of arrays (with the ``.files`` attribute), and for the arrays themselves.
Examples -------- >>> test_array = np.random.rand(3, 2) >>> test_vector = np.random.rand(4) >>> np.savez_compressed('/tmp/123', a=test_array, b=test_vector) >>> loaded = np.load('/tmp/123.npz') >>> print(np.array_equal(test_array, loaded['a'])) True >>> print(np.array_equal(test_vector, loaded['b'])) True
"""
_savez(file, args, kwds, True)
def _savez(file, args, kwds, compress, allow_pickle=True, pickle_kwargs=None): # Import is postponed to here since zipfile depends on gzip, an optional # component of the so-called standard library. import zipfile
if not hasattr(file, 'write'): file = os_fspath(file) if not file.endswith('.npz'): file = file + '.npz'
namedict = kwds for i, val in enumerate(args): key = 'arr_%d' % i if key in namedict.keys(): raise ValueError( "Cannot use un-named variables and keyword %s" % key) namedict[key] = val
if compress: compression = zipfile.ZIP_DEFLATED else: compression = zipfile.ZIP_STORED
zipf = zipfile_factory(file, mode="w", compression=compression)
for key, val in namedict.items(): fname = key + '.npy' val = np.asanyarray(val) # always force zip64, gh-10776 with zipf.open(fname, 'w', force_zip64=True) as fid: format.write_array(fid, val, allow_pickle=allow_pickle, pickle_kwargs=pickle_kwargs)
zipf.close()
def _getconv(dtype): """ Find the correct dtype converter. Adapted from matplotlib """
def floatconv(x): x.lower() if '0x' in x: return float.fromhex(x) return float(x)
typ = dtype.type if issubclass(typ, np.bool_): return lambda x: bool(int(x)) if issubclass(typ, np.uint64): return np.uint64 if issubclass(typ, np.int64): return np.int64 if issubclass(typ, np.integer): return lambda x: int(float(x)) elif issubclass(typ, np.longdouble): return np.longdouble elif issubclass(typ, np.floating): return floatconv elif issubclass(typ, complex): return lambda x: complex(asstr(x).replace('+-', '-')) elif issubclass(typ, np.bytes_): return asbytes elif issubclass(typ, np.unicode_): return asunicode else: return asstr
# amount of lines loadtxt reads in one chunk, can be overridden for testing_loadtxt_chunksize = 50000
def _loadtxt_dispatcher(fname, dtype=None, comments=None, delimiter=None, converters=None, skiprows=None, usecols=None, unpack=None, ndmin=None, encoding=None, max_rows=None, *, like=None): return (like,)
@set_array_function_like_doc@set_module('numpy')def loadtxt(fname, dtype=float, comments='#', delimiter=None, converters=None, skiprows=0, usecols=None, unpack=False, ndmin=0, encoding='bytes', max_rows=None, *, like=None): r"""
Load data from a text file.
Each row in the text file must have the same number of values.
Parameters ---------- fname : file, str, or pathlib.Path File, filename, or generator to read. If the filename extension is ``.gz`` or ``.bz2``, the file is first decompressed. Note that generators should return byte strings. dtype : data-type, optional Data-type of the resulting array; default: float. If this is a structured data-type, the resulting array will be 1-dimensional, and each row will be interpreted as an element of the array. In this case, the number of columns used must match the number of fields in the data-type. comments : str or sequence of str, optional The characters or list of characters used to indicate the start of a comment. None implies no comments. For backwards compatibility, byte strings will be decoded as 'latin1'. The default is '#'. delimiter : str, optional The string used to separate values. For backwards compatibility, byte strings will be decoded as 'latin1'. The default is whitespace. converters : dict, optional A dictionary mapping column number to a function that will parse the column string into the desired value. E.g., if column 0 is a date string: ``converters = {0: datestr2num}``. Converters can also be used to provide a default value for missing data (but see also `genfromtxt`): ``converters = {3: lambda s: float(s.strip() or 0)}``. Default: None. skiprows : int, optional Skip the first `skiprows` lines, including comments; default: 0. usecols : int or sequence, optional Which columns to read, with 0 being the first. For example, ``usecols = (1,4,5)`` will extract the 2nd, 5th and 6th columns. The default, None, results in all columns being read.
.. versionchanged:: 1.11.0 When a single column has to be read it is possible to use an integer instead of a tuple. E.g ``usecols = 3`` reads the fourth column the same way as ``usecols = (3,)`` would. unpack : bool, optional If True, the returned array is transposed, so that arguments may be unpacked using ``x, y, z = loadtxt(...)``. When used with a structured data-type, arrays are returned for each field. Default is False. ndmin : int, optional The returned array will have at least `ndmin` dimensions. Otherwise mono-dimensional axes will be squeezed. Legal values: 0 (default), 1 or 2.
.. versionadded:: 1.6.0 encoding : str, optional Encoding used to decode the inputfile. Does not apply to input streams. The special value 'bytes' enables backward compatibility workarounds that ensures you receive byte arrays as results if possible and passes 'latin1' encoded strings to converters. Override this value to receive unicode arrays and pass strings as input to converters. If set to None the system default is used. The default value is 'bytes'.
.. versionadded:: 1.14.0 max_rows : int, optional Read `max_rows` lines of content after `skiprows` lines. The default is to read all the lines.
.. versionadded:: 1.16.0 ${ARRAY_FUNCTION_LIKE}
.. versionadded:: 1.20.0
Returns ------- out : ndarray Data read from the text file.
See Also -------- load, fromstring, fromregex genfromtxt : Load data with missing values handled as specified. scipy.io.loadmat : reads MATLAB data files
Notes ----- This function aims to be a fast reader for simply formatted files. The `genfromtxt` function provides more sophisticated handling of, e.g., lines with missing values.
.. versionadded:: 1.10.0
The strings produced by the Python float.hex method can be used as input for floats.
Examples -------- >>> from io import StringIO # StringIO behaves like a file object >>> c = StringIO("0 1\n2 3") >>> np.loadtxt(c) array([[0., 1.], [2., 3.]])
>>> d = StringIO("M 21 72\nF 35 58") >>> np.loadtxt(d, dtype={'names': ('gender', 'age', 'weight'), ... 'formats': ('S1', 'i4', 'f4')}) array([(b'M', 21, 72.), (b'F', 35, 58.)], dtype=[('gender', 'S1'), ('age', '<i4'), ('weight', '<f4')])
>>> c = StringIO("1,0,2\n3,0,4") >>> x, y = np.loadtxt(c, delimiter=',', usecols=(0, 2), unpack=True) >>> x array([1., 3.]) >>> y array([2., 4.])
This example shows how `converters` can be used to convert a field with a trailing minus sign into a negative number.
>>> s = StringIO('10.01 31.25-\n19.22 64.31\n17.57- 63.94') >>> def conv(fld): ... return -float(fld[:-1]) if fld.endswith(b'-') else float(fld) ... >>> np.loadtxt(s, converters={0: conv, 1: conv}) array([[ 10.01, -31.25], [ 19.22, 64.31], [-17.57, 63.94]]) """
if like is not None: return _loadtxt_with_like( fname, dtype=dtype, comments=comments, delimiter=delimiter, converters=converters, skiprows=skiprows, usecols=usecols, unpack=unpack, ndmin=ndmin, encoding=encoding, max_rows=max_rows, like=like )
# - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - # Nested functions used by loadtxt. # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
# not to be confused with the flatten_dtype we import... @recursive def flatten_dtype_internal(self, dt): """Unpack a structured data-type, and produce re-packing info.""" if dt.names is None: # If the dtype is flattened, return. # If the dtype has a shape, the dtype occurs # in the list more than once. shape = dt.shape if len(shape) == 0: return ([dt.base], None) else: packing = [(shape[-1], list)] if len(shape) > 1: for dim in dt.shape[-2::-1]: packing = [(dim*packing[0][0], packing*dim)] return ([dt.base] * int(np.prod(dt.shape)), packing) else: types = [] packing = [] for field in dt.names: tp, bytes = dt.fields[field] flat_dt, flat_packing = self(tp) types.extend(flat_dt) # Avoid extra nesting for subarrays if tp.ndim > 0: packing.extend(flat_packing) else: packing.append((len(flat_dt), flat_packing)) return (types, packing)
@recursive def pack_items(self, items, packing): """Pack items into nested lists based on re-packing info.""" if packing is None: return items[0] elif packing is tuple: return tuple(items) elif packing is list: return list(items) else: start = 0 ret = [] for length, subpacking in packing: ret.append(self(items[start:start+length], subpacking)) start += length return tuple(ret)
def split_line(line): """Chop off comments, strip, and split at delimiter. """ line = _decode_line(line, encoding=encoding)
if comments is not None: line = regex_comments.split(line, maxsplit=1)[0] line = line.strip('\r\n') return line.split(delimiter) if line else []
def read_data(chunk_size): """Parse each line, including the first.
The file read, `fh`, is a global defined above.
Parameters ---------- chunk_size : int At most `chunk_size` lines are read at a time, with iteration until all lines are read.
"""
X = [] line_iter = itertools.chain([first_line], fh) line_iter = itertools.islice(line_iter, max_rows) for i, line in enumerate(line_iter): vals = split_line(line) if len(vals) == 0: continue if usecols: vals = [vals[j] for j in usecols] if len(vals) != N: line_num = i + skiprows + 1 raise ValueError("Wrong number of columns at line %d" % line_num)
# Convert each value according to its column and store items = [conv(val) for (conv, val) in zip(converters, vals)]
# Then pack it according to the dtype's nesting items = pack_items(items, packing) X.append(items) if len(X) > chunk_size: yield X X = [] if X: yield X
# - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - # Main body of loadtxt. # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
# Check correctness of the values of `ndmin` if ndmin not in [0, 1, 2]: raise ValueError('Illegal value of ndmin keyword: %s' % ndmin)
# Type conversions for Py3 convenience if comments is not None: if isinstance(comments, (str, bytes)): comments = [comments] comments = [_decode_line(x) for x in comments] # Compile regex for comments beforehand comments = (re.escape(comment) for comment in comments) regex_comments = re.compile('|'.join(comments))
if delimiter is not None: delimiter = _decode_line(delimiter)
user_converters = converters
byte_converters = False if encoding == 'bytes': encoding = None byte_converters = True
if usecols is not None: # Allow usecols to be a single int or a sequence of ints try: usecols_as_list = list(usecols) except TypeError: usecols_as_list = [usecols] for col_idx in usecols_as_list: try: opindex(col_idx) except TypeError as e: e.args = ( "usecols must be an int or a sequence of ints but " "it contains at least one element of type %s" % type(col_idx), ) raise # Fall back to existing code usecols = usecols_as_list
# Make sure we're dealing with a proper dtype dtype = np.dtype(dtype) defconv = _getconv(dtype)
dtype_types, packing = flatten_dtype_internal(dtype)
fown = False try: if isinstance(fname, os_PathLike): fname = os_fspath(fname) if _is_string_like(fname): fh = np.lib._datasource.open(fname, 'rt', encoding=encoding) fencoding = getattr(fh, 'encoding', 'latin1') fh = iter(fh) fown = True else: fh = iter(fname) fencoding = getattr(fname, 'encoding', 'latin1') except TypeError as e: raise ValueError( 'fname must be a string, file handle, or generator' ) from e
# input may be a python2 io stream if encoding is not None: fencoding = encoding # we must assume local encoding # TODO emit portability warning? elif fencoding is None: import locale fencoding = locale.getpreferredencoding()
try: # Skip the first `skiprows` lines for i in range(skiprows): next(fh)
# Read until we find a line with some values, and use # it to estimate the number of columns, N. first_vals = None try: while not first_vals: first_line = next(fh) first_vals = split_line(first_line) except StopIteration: # End of lines reached first_line = '' first_vals = [] warnings.warn('loadtxt: Empty input file: "%s"' % fname, stacklevel=2) N = len(usecols or first_vals)
# Now that we know N, create the default converters list, and # set packing, if necessary. if len(dtype_types) > 1: # We're dealing with a structured array, each field of # the dtype matches a column converters = [_getconv(dt) for dt in dtype_types] else: # All fields have the same dtype converters = [defconv for i in range(N)] if N > 1: packing = [(N, tuple)]
# By preference, use the converters specified by the user for i, conv in (user_converters or {}).items(): if usecols: try: i = usecols.index(i) except ValueError: # Unused converter specified continue if byte_converters: # converters may use decode to workaround numpy's old # behaviour, so encode the string again before passing to # the user converter def tobytes_first(x, conv): if type(x) is bytes: return conv(x) return conv(x.encode("latin1")) converters[i] = functools.partial(tobytes_first, conv=conv) else: converters[i] = conv
converters = [conv if conv is not bytes else lambda x: x.encode(fencoding) for conv in converters]
# read data in chunks and fill it into an array via resize # over-allocating and shrinking the array later may be faster but is # probably not relevant compared to the cost of actually reading and # converting the data X = None for x in read_data(_loadtxt_chunksize): if X is None: X = np.array(x, dtype) else: nshape = list(X.shape) pos = nshape[0] nshape[0] += len(x) X.resize(nshape, refcheck=False) X[pos:, ...] = x finally: if fown: fh.close()
if X is None: X = np.array([], dtype)
# Multicolumn data are returned with shape (1, N, M), i.e. # (1, 1, M) for a single row - remove the singleton dimension there if X.ndim == 3 and X.shape[:2] == (1, 1): X.shape = (1, -1)
# Verify that the array has at least dimensions `ndmin`. # Tweak the size and shape of the arrays - remove extraneous dimensions if X.ndim > ndmin: X = np.squeeze(X) # and ensure we have the minimum number of dimensions asked for # - has to be in this order for the odd case ndmin=1, X.squeeze().ndim=0 if X.ndim < ndmin: if ndmin == 1: X = np.atleast_1d(X) elif ndmin == 2: X = np.atleast_2d(X).T
if unpack: if len(dtype_types) > 1: # For structured arrays, return an array for each field. return [X[field] for field in dtype.names] else: return X.T else: return X
_loadtxt_with_like = array_function_dispatch( _loadtxt_dispatcher)(loadtxt)
def _savetxt_dispatcher(fname, X, fmt=None, delimiter=None, newline=None, header=None, footer=None, comments=None, encoding=None): return (X,)
@array_function_dispatch(_savetxt_dispatcher)def savetxt(fname, X, fmt='%.18e', delimiter=' ', newline='\n', header='', footer='', comments='# ', encoding=None): """
Save an array to a text file.
Parameters ---------- fname : filename or file handle If the filename ends in ``.gz``, the file is automatically saved in compressed gzip format. `loadtxt` understands gzipped files transparently. X : 1D or 2D array_like Data to be saved to a text file. fmt : str or sequence of strs, optional A single format (%10.5f), a sequence of formats, or a multi-format string, e.g. 'Iteration %d -- %10.5f', in which case `delimiter` is ignored. For complex `X`, the legal options for `fmt` are:
* a single specifier, `fmt='%.4e'`, resulting in numbers formatted like `' (%s+%sj)' % (fmt, fmt)` * a full string specifying every real and imaginary part, e.g. `' %.4e %+.4ej %.4e %+.4ej %.4e %+.4ej'` for 3 columns * a list of specifiers, one per column - in this case, the real and imaginary part must have separate specifiers, e.g. `['%.3e + %.3ej', '(%.15e%+.15ej)']` for 2 columns delimiter : str, optional String or character separating columns. newline : str, optional String or character separating lines.
.. versionadded:: 1.5.0 header : str, optional String that will be written at the beginning of the file.
.. versionadded:: 1.7.0 footer : str, optional String that will be written at the end of the file.
.. versionadded:: 1.7.0 comments : str, optional String that will be prepended to the ``header`` and ``footer`` strings, to mark them as comments. Default: '# ', as expected by e.g. ``numpy.loadtxt``.
.. versionadded:: 1.7.0 encoding : {None, str}, optional Encoding used to encode the outputfile. Does not apply to output streams. If the encoding is something other than 'bytes' or 'latin1' you will not be able to load the file in NumPy versions < 1.14. Default is 'latin1'.
.. versionadded:: 1.14.0
See Also -------- save : Save an array to a binary file in NumPy ``.npy`` format savez : Save several arrays into an uncompressed ``.npz`` archive savez_compressed : Save several arrays into a compressed ``.npz`` archive
Notes ----- Further explanation of the `fmt` parameter (``%[flag]width[.precision]specifier``):
flags: ``-`` : left justify
``+`` : Forces to precede result with + or -.
``0`` : Left pad the number with zeros instead of space (see width).
width: Minimum number of characters to be printed. The value is not truncated if it has more characters.
precision: - For integer specifiers (eg. ``d,i,o,x``), the minimum number of digits. - For ``e, E`` and ``f`` specifiers, the number of digits to print after the decimal point. - For ``g`` and ``G``, the maximum number of significant digits. - For ``s``, the maximum number of characters.
specifiers: ``c`` : character
``d`` or ``i`` : signed decimal integer
``e`` or ``E`` : scientific notation with ``e`` or ``E``.
``f`` : decimal floating point
``g,G`` : use the shorter of ``e,E`` or ``f``
``o`` : signed octal
``s`` : string of characters
``u`` : unsigned decimal integer
``x,X`` : unsigned hexadecimal integer
This explanation of ``fmt`` is not complete, for an exhaustive specification see [1]_.
References ---------- .. [1] `Format Specification Mini-Language <https://docs.python.org/library/string.html#format-specification-mini-language>`_, Python Documentation.
Examples -------- >>> x = y = z = np.arange(0.0,5.0,1.0) >>> np.savetxt('test.out', x, delimiter=',') # X is an array >>> np.savetxt('test.out', (x,y,z)) # x,y,z equal sized 1D arrays >>> np.savetxt('test.out', x, fmt='%1.4e') # use exponential notation
"""
# Py3 conversions first if isinstance(fmt, bytes): fmt = asstr(fmt) delimiter = asstr(delimiter)
class WriteWrap: """Convert to bytes on bytestream inputs.
"""
def __init__(self, fh, encoding): self.fh = fh self.encoding = encoding self.do_write = self.first_write
def close(self): self.fh.close()
def write(self, v): self.do_write(v)
def write_bytes(self, v): if isinstance(v, bytes): self.fh.write(v) else: self.fh.write(v.encode(self.encoding))
def write_normal(self, v): self.fh.write(asunicode(v))
def first_write(self, v): try: self.write_normal(v) self.write = self.write_normal except TypeError: # input is probably a bytestream self.write_bytes(v) self.write = self.write_bytes
own_fh = False if isinstance(fname, os_PathLike): fname = os_fspath(fname) if _is_string_like(fname): # datasource doesn't support creating a new file ... open(fname, 'wt').close() fh = np.lib._datasource.open(fname, 'wt', encoding=encoding) own_fh = True elif hasattr(fname, 'write'): # wrap to handle byte output streams fh = WriteWrap(fname, encoding or 'latin1') else: raise ValueError('fname must be a string or file handle')
try: X = np.asarray(X)
# Handle 1-dimensional arrays if X.ndim == 0 or X.ndim > 2: raise ValueError( "Expected 1D or 2D array, got %dD array instead" % X.ndim) elif X.ndim == 1: # Common case -- 1d array of numbers if X.dtype.names is None: X = np.atleast_2d(X).T ncol = 1
# Complex dtype -- each field indicates a separate column else: ncol = len(X.dtype.names) else: ncol = X.shape[1]
iscomplex_X = np.iscomplexobj(X) # `fmt` can be a string with multiple insertion points or a # list of formats. E.g. '%10.5f\t%10d' or ('%10.5f', '$10d') if type(fmt) in (list, tuple): if len(fmt) != ncol: raise AttributeError('fmt has wrong shape. %s' % str(fmt)) format = asstr(delimiter).join(map(asstr, fmt)) elif isinstance(fmt, str): n_fmt_chars = fmt.count('%') error = ValueError('fmt has wrong number of %% formats: %s' % fmt) if n_fmt_chars == 1: if iscomplex_X: fmt = [' (%s+%sj)' % (fmt, fmt), ] * ncol else: fmt = [fmt, ] * ncol format = delimiter.join(fmt) elif iscomplex_X and n_fmt_chars != (2 * ncol): raise error elif ((not iscomplex_X) and n_fmt_chars != ncol): raise error else: format = fmt else: raise ValueError('invalid fmt: %r' % (fmt,))
if len(header) > 0: header = header.replace('\n', '\n' + comments) fh.write(comments + header + newline) if iscomplex_X: for row in X: row2 = [] for number in row: row2.append(number.real) row2.append(number.imag) s = format % tuple(row2) + newline fh.write(s.replace('+-', '-')) else: for row in X: try: v = format % tuple(row) + newline except TypeError as e: raise TypeError("Mismatch between array dtype ('%s') and " "format specifier ('%s')" % (str(X.dtype), format)) from e fh.write(v)
if len(footer) > 0: footer = footer.replace('\n', '\n' + comments) fh.write(comments + footer + newline) finally: if own_fh: fh.close()
@set_module('numpy')def fromregex(file, regexp, dtype, encoding=None): """
Construct an array from a text file, using regular expression parsing.
The returned array is always a structured array, and is constructed from all matches of the regular expression in the file. Groups in the regular expression are converted to fields of the structured array.
Parameters ---------- file : str or file Filename or file object to read. regexp : str or regexp Regular expression used to parse the file. Groups in the regular expression correspond to fields in the dtype. dtype : dtype or list of dtypes Dtype for the structured array. encoding : str, optional Encoding used to decode the inputfile. Does not apply to input streams.
.. versionadded:: 1.14.0
Returns ------- output : ndarray The output array, containing the part of the content of `file` that was matched by `regexp`. `output` is always a structured array.
Raises ------ TypeError When `dtype` is not a valid dtype for a structured array.
See Also -------- fromstring, loadtxt
Notes ----- Dtypes for structured arrays can be specified in several forms, but all forms specify at least the data type and field name. For details see `basics.rec`.
Examples -------- >>> f = open('test.dat', 'w') >>> _ = f.write("1312 foo\\n1534 bar\\n444 qux") >>> f.close()
>>> regexp = r"(\\d+)\\s+(...)" # match [digits, whitespace, anything] >>> output = np.fromregex('test.dat', regexp, ... [('num', np.int64), ('key', 'S3')]) >>> output array([(1312, b'foo'), (1534, b'bar'), ( 444, b'qux')], dtype=[('num', '<i8'), ('key', 'S3')]) >>> output['num'] array([1312, 1534, 444])
"""
own_fh = False if not hasattr(file, "read"): file = np.lib._datasource.open(file, 'rt', encoding=encoding) own_fh = True
try: if not isinstance(dtype, np.dtype): dtype = np.dtype(dtype)
content = file.read() if isinstance(content, bytes) and isinstance(regexp, np.compat.unicode): regexp = asbytes(regexp) elif isinstance(content, np.compat.unicode) and isinstance(regexp, bytes): regexp = asstr(regexp)
if not hasattr(regexp, 'match'): regexp = re.compile(regexp) seq = regexp.findall(content) if seq and not isinstance(seq[0], tuple): # Only one group is in the regexp. # Create the new array as a single data-type and then # re-interpret as a single-field structured array. newdtype = np.dtype(dtype[dtype.names[0]]) output = np.array(seq, dtype=newdtype) output.dtype = dtype else: output = np.array(seq, dtype=dtype)
return output finally: if own_fh: file.close()
#####--------------------------------------------------------------------------#---- --- ASCII functions ---#####--------------------------------------------------------------------------
def _genfromtxt_dispatcher(fname, dtype=None, comments=None, delimiter=None, skip_header=None, skip_footer=None, converters=None, missing_values=None, filling_values=None, usecols=None, names=None, excludelist=None, deletechars=None, replace_space=None, autostrip=None, case_sensitive=None, defaultfmt=None, unpack=None, usemask=None, loose=None, invalid_raise=None, max_rows=None, encoding=None, *, like=None): return (like,)
@set_array_function_like_doc@set_module('numpy')def genfromtxt(fname, dtype=float, comments='#', delimiter=None, skip_header=0, skip_footer=0, converters=None, missing_values=None, filling_values=None, usecols=None, names=None, excludelist=None, deletechars=''.join(sorted(NameValidator.defaultdeletechars)), replace_space='_', autostrip=False, case_sensitive=True, defaultfmt="f%i", unpack=None, usemask=False, loose=True, invalid_raise=True, max_rows=None, encoding='bytes', *, like=None): """
Load data from a text file, with missing values handled as specified.
Each line past the first `skip_header` lines is split at the `delimiter` character, and characters following the `comments` character are discarded.
Parameters ---------- fname : file, str, pathlib.Path, list of str, generator File, filename, list, or generator to read. If the filename extension is `.gz` or `.bz2`, the file is first decompressed. Note that generators must return byte strings. The strings in a list or produced by a generator are treated as lines. dtype : dtype, optional Data type of the resulting array. If None, the dtypes will be determined by the contents of each column, individually. comments : str, optional The character used to indicate the start of a comment. All the characters occurring on a line after a comment are discarded. delimiter : str, int, or sequence, optional The string used to separate values. By default, any consecutive whitespaces act as delimiter. An integer or sequence of integers can also be provided as width(s) of each field. skiprows : int, optional `skiprows` was removed in numpy 1.10. Please use `skip_header` instead. skip_header : int, optional The number of lines to skip at the beginning of the file. skip_footer : int, optional The number of lines to skip at the end of the file. converters : variable, optional The set of functions that convert the data of a column to a value. The converters can also be used to provide a default value for missing data: ``converters = {3: lambda s: float(s or 0)}``. missing : variable, optional `missing` was removed in numpy 1.10. Please use `missing_values` instead. missing_values : variable, optional The set of strings corresponding to missing data. filling_values : variable, optional The set of values to be used as default when the data are missing. usecols : sequence, optional Which columns to read, with 0 being the first. For example, ``usecols = (1, 4, 5)`` will extract the 2nd, 5th and 6th columns. names : {None, True, str, sequence}, optional If `names` is True, the field names are read from the first line after the first `skip_header` lines. This line can optionally be preceeded by a comment delimiter. If `names` is a sequence or a single-string of comma-separated names, the names will be used to define the field names in a structured dtype. If `names` is None, the names of the dtype fields will be used, if any. excludelist : sequence, optional A list of names to exclude. This list is appended to the default list ['return','file','print']. Excluded names are appended with an underscore: for example, `file` would become `file_`. deletechars : str, optional A string combining invalid characters that must be deleted from the names. defaultfmt : str, optional A format used to define default field names, such as "f%i" or "f_%02i". autostrip : bool, optional Whether to automatically strip white spaces from the variables. replace_space : char, optional Character(s) used in replacement of white spaces in the variable names. By default, use a '_'. case_sensitive : {True, False, 'upper', 'lower'}, optional If True, field names are case sensitive. If False or 'upper', field names are converted to upper case. If 'lower', field names are converted to lower case. unpack : bool, optional If True, the returned array is transposed, so that arguments may be unpacked using ``x, y, z = genfromtxt(...)``. When used with a structured data-type, arrays are returned for each field. Default is False. usemask : bool, optional If True, return a masked array. If False, return a regular array. loose : bool, optional If True, do not raise errors for invalid values. invalid_raise : bool, optional If True, an exception is raised if an inconsistency is detected in the number of columns. If False, a warning is emitted and the offending lines are skipped. max_rows : int, optional The maximum number of rows to read. Must not be used with skip_footer at the same time. If given, the value must be at least 1. Default is to read the entire file.
.. versionadded:: 1.10.0 encoding : str, optional Encoding used to decode the inputfile. Does not apply when `fname` is a file object. The special value 'bytes' enables backward compatibility workarounds that ensure that you receive byte arrays when possible and passes latin1 encoded strings to converters. Override this value to receive unicode arrays and pass strings as input to converters. If set to None the system default is used. The default value is 'bytes'.
.. versionadded:: 1.14.0 ${ARRAY_FUNCTION_LIKE}
.. versionadded:: 1.20.0
Returns ------- out : ndarray Data read from the text file. If `usemask` is True, this is a masked array.
See Also -------- numpy.loadtxt : equivalent function when no data is missing.
Notes ----- * When spaces are used as delimiters, or when no delimiter has been given as input, there should not be any missing data between two fields. * When the variables are named (either by a flexible dtype or with `names`), there must not be any header in the file (else a ValueError exception is raised). * Individual values are not stripped of spaces by default. When using a custom converter, make sure the function does remove spaces.
References ---------- .. [1] NumPy User Guide, section `I/O with NumPy <https://docs.scipy.org/doc/numpy/user/basics.io.genfromtxt.html>`_.
Examples -------- >>> from io import StringIO >>> import numpy as np
Comma delimited file with mixed dtype
>>> s = StringIO(u"1,1.3,abcde") >>> data = np.genfromtxt(s, dtype=[('myint','i8'),('myfloat','f8'), ... ('mystring','S5')], delimiter=",") >>> data array((1, 1.3, b'abcde'), dtype=[('myint', '<i8'), ('myfloat', '<f8'), ('mystring', 'S5')])
Using dtype = None
>>> _ = s.seek(0) # needed for StringIO example only >>> data = np.genfromtxt(s, dtype=None, ... names = ['myint','myfloat','mystring'], delimiter=",") >>> data array((1, 1.3, b'abcde'), dtype=[('myint', '<i8'), ('myfloat', '<f8'), ('mystring', 'S5')])
Specifying dtype and names
>>> _ = s.seek(0) >>> data = np.genfromtxt(s, dtype="i8,f8,S5", ... names=['myint','myfloat','mystring'], delimiter=",") >>> data array((1, 1.3, b'abcde'), dtype=[('myint', '<i8'), ('myfloat', '<f8'), ('mystring', 'S5')])
An example with fixed-width columns
>>> s = StringIO(u"11.3abcde") >>> data = np.genfromtxt(s, dtype=None, names=['intvar','fltvar','strvar'], ... delimiter=[1,3,5]) >>> data array((1, 1.3, b'abcde'), dtype=[('intvar', '<i8'), ('fltvar', '<f8'), ('strvar', 'S5')])
An example to show comments
>>> f = StringIO('''
... text,# of chars ... hello world,11 ... numpy,5''')
>>> np.genfromtxt(f, dtype='S12,S12', delimiter=',') array([(b'text', b''), (b'hello world', b'11'), (b'numpy', b'5')], dtype=[('f0', 'S12'), ('f1', 'S12')])
"""
if like is not None: return _genfromtxt_with_like( fname, dtype=dtype, comments=comments, delimiter=delimiter, skip_header=skip_header, skip_footer=skip_footer, converters=converters, missing_values=missing_values, filling_values=filling_values, usecols=usecols, names=names, excludelist=excludelist, deletechars=deletechars, replace_space=replace_space, autostrip=autostrip, case_sensitive=case_sensitive, defaultfmt=defaultfmt, unpack=unpack, usemask=usemask, loose=loose, invalid_raise=invalid_raise, max_rows=max_rows, encoding=encoding, like=like )
if max_rows is not None: if skip_footer: raise ValueError( "The keywords 'skip_footer' and 'max_rows' can not be " "specified at the same time.") if max_rows < 1: raise ValueError("'max_rows' must be at least 1.")
if usemask: from numpy.ma import MaskedArray, make_mask_descr # Check the input dictionary of converters user_converters = converters or {} if not isinstance(user_converters, dict): raise TypeError( "The input argument 'converter' should be a valid dictionary " "(got '%s' instead)" % type(user_converters))
if encoding == 'bytes': encoding = None byte_converters = True else: byte_converters = False
# Initialize the filehandle, the LineSplitter and the NameValidator try: if isinstance(fname, os_PathLike): fname = os_fspath(fname) if isinstance(fname, str): fid = np.lib._datasource.open(fname, 'rt', encoding=encoding) fid_ctx = contextlib.closing(fid) else: fid = fname fid_ctx = contextlib.nullcontext(fid) fhd = iter(fid) except TypeError as e: raise TypeError( "fname must be a string, filehandle, list of strings, " "or generator. Got %s instead." % type(fname)) from e
with fid_ctx: split_line = LineSplitter(delimiter=delimiter, comments=comments, autostrip=autostrip, encoding=encoding) validate_names = NameValidator(excludelist=excludelist, deletechars=deletechars, case_sensitive=case_sensitive, replace_space=replace_space)
# Skip the first `skip_header` rows try: for i in range(skip_header): next(fhd)
# Keep on until we find the first valid values first_values = None
while not first_values: first_line = _decode_line(next(fhd), encoding) if (names is True) and (comments is not None): if comments in first_line: first_line = ( ''.join(first_line.split(comments)[1:])) first_values = split_line(first_line) except StopIteration: # return an empty array if the datafile is empty first_line = '' first_values = [] warnings.warn('genfromtxt: Empty input file: "%s"' % fname, stacklevel=2)
# Should we take the first values as names ? if names is True: fval = first_values[0].strip() if comments is not None: if fval in comments: del first_values[0]
# Check the columns to use: make sure `usecols` is a list if usecols is not None: try: usecols = [_.strip() for _ in usecols.split(",")] except AttributeError: try: usecols = list(usecols) except TypeError: usecols = [usecols, ] nbcols = len(usecols or first_values)
# Check the names and overwrite the dtype.names if needed if names is True: names = validate_names([str(_.strip()) for _ in first_values]) first_line = '' elif _is_string_like(names): names = validate_names([_.strip() for _ in names.split(',')]) elif names: names = validate_names(names) # Get the dtype if dtype is not None: dtype = easy_dtype(dtype, defaultfmt=defaultfmt, names=names, excludelist=excludelist, deletechars=deletechars, case_sensitive=case_sensitive, replace_space=replace_space) # Make sure the names is a list (for 2.5) if names is not None: names = list(names)
if usecols: for (i, current) in enumerate(usecols): # if usecols is a list of names, convert to a list of indices if _is_string_like(current): usecols[i] = names.index(current) elif current < 0: usecols[i] = current + len(first_values) # If the dtype is not None, make sure we update it if (dtype is not None) and (len(dtype) > nbcols): descr = dtype.descr dtype = np.dtype([descr[_] for _ in usecols]) names = list(dtype.names) # If `names` is not None, update the names elif (names is not None) and (len(names) > nbcols): names = [names[_] for _ in usecols] elif (names is not None) and (dtype is not None): names = list(dtype.names)
# Process the missing values ............................... # Rename missing_values for convenience user_missing_values = missing_values or () if isinstance(user_missing_values, bytes): user_missing_values = user_missing_values.decode('latin1')
# Define the list of missing_values (one column: one list) missing_values = [list(['']) for _ in range(nbcols)]
# We have a dictionary: process it field by field if isinstance(user_missing_values, dict): # Loop on the items for (key, val) in user_missing_values.items(): # Is the key a string ? if _is_string_like(key): try: # Transform it into an integer key = names.index(key) except ValueError: # We couldn't find it: the name must have been dropped continue # Redefine the key as needed if it's a column number if usecols: try: key = usecols.index(key) except ValueError: pass # Transform the value as a list of string if isinstance(val, (list, tuple)): val = [str(_) for _ in val] else: val = [str(val), ] # Add the value(s) to the current list of missing if key is None: # None acts as default for miss in missing_values: miss.extend(val) else: missing_values[key].extend(val) # We have a sequence : each item matches a column elif isinstance(user_missing_values, (list, tuple)): for (value, entry) in zip(user_missing_values, missing_values): value = str(value) if value not in entry: entry.append(value) # We have a string : apply it to all entries elif isinstance(user_missing_values, str): user_value = user_missing_values.split(",") for entry in missing_values: entry.extend(user_value) # We have something else: apply it to all entries else: for entry in missing_values: entry.extend([str(user_missing_values)])
# Process the filling_values ............................... # Rename the input for convenience user_filling_values = filling_values if user_filling_values is None: user_filling_values = [] # Define the default filling_values = [None] * nbcols # We have a dictionary : update each entry individually if isinstance(user_filling_values, dict): for (key, val) in user_filling_values.items(): if _is_string_like(key): try: # Transform it into an integer key = names.index(key) except ValueError: # We couldn't find it: the name must have been dropped, continue # Redefine the key if it's a column number and usecols is defined if usecols: try: key = usecols.index(key) except ValueError: pass # Add the value to the list filling_values[key] = val # We have a sequence : update on a one-to-one basis elif isinstance(user_filling_values, (list, tuple)): n = len(user_filling_values) if (n <= nbcols): filling_values[:n] = user_filling_values else: filling_values = user_filling_values[:nbcols] # We have something else : use it for all entries else: filling_values = [user_filling_values] * nbcols
# Initialize the converters ................................ if dtype is None: # Note: we can't use a [...]*nbcols, as we would have 3 times the same # ... converter, instead of 3 different converters. converters = [StringConverter(None, missing_values=miss, default=fill) for (miss, fill) in zip(missing_values, filling_values)] else: dtype_flat = flatten_dtype(dtype, flatten_base=True) # Initialize the converters if len(dtype_flat) > 1: # Flexible type : get a converter from each dtype zipit = zip(dtype_flat, missing_values, filling_values) converters = [StringConverter(dt, locked=True, missing_values=miss, default=fill) for (dt, miss, fill) in zipit] else: # Set to a default converter (but w/ different missing values) zipit = zip(missing_values, filling_values) converters = [StringConverter(dtype, locked=True, missing_values=miss, default=fill) for (miss, fill) in zipit] # Update the converters to use the user-defined ones uc_update = [] for (j, conv) in user_converters.items(): # If the converter is specified by column names, use the index instead if _is_string_like(j): try: j = names.index(j) i = j except ValueError: continue elif usecols: try: i = usecols.index(j) except ValueError: # Unused converter specified continue else: i = j # Find the value to test - first_line is not filtered by usecols: if len(first_line): testing_value = first_values[j] else: testing_value = None if conv is bytes: user_conv = asbytes elif byte_converters: # converters may use decode to workaround numpy's old behaviour, # so encode the string again before passing to the user converter def tobytes_first(x, conv): if type(x) is bytes: return conv(x) return conv(x.encode("latin1")) user_conv = functools.partial(tobytes_first, conv=conv) else: user_conv = conv converters[i].update(user_conv, locked=True, testing_value=testing_value, default=filling_values[i], missing_values=missing_values[i],) uc_update.append((i, user_conv)) # Make sure we have the corrected keys in user_converters... user_converters.update(uc_update)
# Fixme: possible error as following variable never used. # miss_chars = [_.missing_values for _ in converters]
# Initialize the output lists ... # ... rows rows = [] append_to_rows = rows.append # ... masks if usemask: masks = [] append_to_masks = masks.append # ... invalid invalid = [] append_to_invalid = invalid.append
# Parse each line for (i, line) in enumerate(itertools.chain([first_line, ], fhd)): values = split_line(line) nbvalues = len(values) # Skip an empty line if nbvalues == 0: continue if usecols: # Select only the columns we need try: values = [values[_] for _ in usecols] except IndexError: append_to_invalid((i + skip_header + 1, nbvalues)) continue elif nbvalues != nbcols: append_to_invalid((i + skip_header + 1, nbvalues)) continue # Store the values append_to_rows(tuple(values)) if usemask: append_to_masks(tuple([v.strip() in m for (v, m) in zip(values, missing_values)])) if len(rows) == max_rows: break
# Upgrade the converters (if needed) if dtype is None: for (i, converter) in enumerate(converters): current_column = [itemgetter(i)(_m) for _m in rows] try: converter.iterupgrade(current_column) except ConverterLockError: errmsg = "Converter #%i is locked and cannot be upgraded: " % i current_column = map(itemgetter(i), rows) for (j, value) in enumerate(current_column): try: converter.upgrade(value) except (ConverterError, ValueError): errmsg += "(occurred line #%i for value '%s')" errmsg %= (j + 1 + skip_header, value) raise ConverterError(errmsg)
# Check that we don't have invalid values nbinvalid = len(invalid) if nbinvalid > 0: nbrows = len(rows) + nbinvalid - skip_footer # Construct the error message template = " Line #%%i (got %%i columns instead of %i)" % nbcols if skip_footer > 0: nbinvalid_skipped = len([_ for _ in invalid if _[0] > nbrows + skip_header]) invalid = invalid[:nbinvalid - nbinvalid_skipped] skip_footer -= nbinvalid_skipped## nbrows -= skip_footer# errmsg = [template % (i, nb)# for (i, nb) in invalid if i < nbrows]# else: errmsg = [template % (i, nb) for (i, nb) in invalid] if len(errmsg): errmsg.insert(0, "Some errors were detected !") errmsg = "\n".join(errmsg) # Raise an exception ? if invalid_raise: raise ValueError(errmsg) # Issue a warning ? else: warnings.warn(errmsg, ConversionWarning, stacklevel=2)
# Strip the last skip_footer data if skip_footer > 0: rows = rows[:-skip_footer] if usemask: masks = masks[:-skip_footer]
# Convert each value according to the converter: # We want to modify the list in place to avoid creating a new one... if loose: rows = list( zip(*[[conv._loose_call(_r) for _r in map(itemgetter(i), rows)] for (i, conv) in enumerate(converters)])) else: rows = list( zip(*[[conv._strict_call(_r) for _r in map(itemgetter(i), rows)] for (i, conv) in enumerate(converters)]))
# Reset the dtype data = rows if dtype is None: # Get the dtypes from the types of the converters column_types = [conv.type for conv in converters] # Find the columns with strings... strcolidx = [i for (i, v) in enumerate(column_types) if v == np.unicode_]
if byte_converters and strcolidx: # convert strings back to bytes for backward compatibility warnings.warn( "Reading unicode strings without specifying the encoding " "argument is deprecated. Set the encoding, use None for the " "system default.", np.VisibleDeprecationWarning, stacklevel=2) def encode_unicode_cols(row_tup): row = list(row_tup) for i in strcolidx: row[i] = row[i].encode('latin1') return tuple(row)
try: data = [encode_unicode_cols(r) for r in data] except UnicodeEncodeError: pass else: for i in strcolidx: column_types[i] = np.bytes_
# Update string types to be the right length sized_column_types = column_types[:] for i, col_type in enumerate(column_types): if np.issubdtype(col_type, np.character): n_chars = max(len(row[i]) for row in data) sized_column_types[i] = (col_type, n_chars)
if names is None: # If the dtype is uniform (before sizing strings) base = { c_type for c, c_type in zip(converters, column_types) if c._checked} if len(base) == 1: uniform_type, = base (ddtype, mdtype) = (uniform_type, bool) else: ddtype = [(defaultfmt % i, dt) for (i, dt) in enumerate(sized_column_types)] if usemask: mdtype = [(defaultfmt % i, bool) for (i, dt) in enumerate(sized_column_types)] else: ddtype = list(zip(names, sized_column_types)) mdtype = list(zip(names, [bool] * len(sized_column_types))) output = np.array(data, dtype=ddtype) if usemask: outputmask = np.array(masks, dtype=mdtype) else: # Overwrite the initial dtype names if needed if names and dtype.names is not None: dtype.names = names # Case 1. We have a structured type if len(dtype_flat) > 1: # Nested dtype, eg [('a', int), ('b', [('b0', int), ('b1', 'f4')])] # First, create the array using a flattened dtype: # [('a', int), ('b1', int), ('b2', float)] # Then, view the array using the specified dtype. if 'O' in (_.char for _ in dtype_flat): if has_nested_fields(dtype): raise NotImplementedError( "Nested fields involving objects are not supported...") else: output = np.array(data, dtype=dtype) else: rows = np.array(data, dtype=[('', _) for _ in dtype_flat]) output = rows.view(dtype) # Now, process the rowmasks the same way if usemask: rowmasks = np.array( masks, dtype=np.dtype([('', bool) for t in dtype_flat])) # Construct the new dtype mdtype = make_mask_descr(dtype) outputmask = rowmasks.view(mdtype) # Case #2. We have a basic dtype else: # We used some user-defined converters if user_converters: ishomogeneous = True descr = [] for i, ttype in enumerate([conv.type for conv in converters]): # Keep the dtype of the current converter if i in user_converters: ishomogeneous &= (ttype == dtype.type) if np.issubdtype(ttype, np.character): ttype = (ttype, max(len(row[i]) for row in data)) descr.append(('', ttype)) else: descr.append(('', dtype)) # So we changed the dtype ? if not ishomogeneous: # We have more than one field if len(descr) > 1: dtype = np.dtype(descr) # We have only one field: drop the name if not needed. else: dtype = np.dtype(ttype) # output = np.array(data, dtype) if usemask: if dtype.names is not None: mdtype = [(_, bool) for _ in dtype.names] else: mdtype = bool outputmask = np.array(masks, dtype=mdtype) # Try to take care of the missing data we missed names = output.dtype.names if usemask and names: for (name, conv) in zip(names, converters): missing_values = [conv(_) for _ in conv.missing_values if _ != ''] for mval in missing_values: outputmask[name] |= (output[name] == mval) # Construct the final array if usemask: output = output.view(MaskedArray) output._mask = outputmask output = np.squeeze(output) if unpack: if names is None: return output.T elif len(names) == 1: # squeeze single-name dtypes too return output[names[0]] else: # For structured arrays with multiple fields, # return an array for each field. return [output[field] for field in names] return output
_genfromtxt_with_like = array_function_dispatch( _genfromtxt_dispatcher)(genfromtxt)
def ndfromtxt(fname, **kwargs): """
Load ASCII data stored in a file and return it as a single array.
.. deprecated:: 1.17 ndfromtxt` is a deprecated alias of `genfromtxt` which overwrites the ``usemask`` argument with `False` even when explicitly called as ``ndfromtxt(..., usemask=True)``. Use `genfromtxt` instead.
Parameters ---------- fname, kwargs : For a description of input parameters, see `genfromtxt`.
See Also -------- numpy.genfromtxt : generic function.
"""
kwargs['usemask'] = False # Numpy 1.17 warnings.warn( "np.ndfromtxt is a deprecated alias of np.genfromtxt, " "prefer the latter.", DeprecationWarning, stacklevel=2) return genfromtxt(fname, **kwargs)
def mafromtxt(fname, **kwargs): """
Load ASCII data stored in a text file and return a masked array.
.. deprecated:: 1.17 np.mafromtxt is a deprecated alias of `genfromtxt` which overwrites the ``usemask`` argument with `True` even when explicitly called as ``mafromtxt(..., usemask=False)``. Use `genfromtxt` instead.
Parameters ---------- fname, kwargs : For a description of input parameters, see `genfromtxt`.
See Also -------- numpy.genfromtxt : generic function to load ASCII data.
"""
kwargs['usemask'] = True # Numpy 1.17 warnings.warn( "np.mafromtxt is a deprecated alias of np.genfromtxt, " "prefer the latter.", DeprecationWarning, stacklevel=2) return genfromtxt(fname, **kwargs)
def recfromtxt(fname, **kwargs): """
Load ASCII data from a file and return it in a record array.
If ``usemask=False`` a standard `recarray` is returned, if ``usemask=True`` a MaskedRecords array is returned.
Parameters ---------- fname, kwargs : For a description of input parameters, see `genfromtxt`.
See Also -------- numpy.genfromtxt : generic function
Notes ----- By default, `dtype` is None, which means that the data-type of the output array will be determined from the data.
"""
kwargs.setdefault("dtype", None) usemask = kwargs.get('usemask', False) output = genfromtxt(fname, **kwargs) if usemask: from numpy.ma.mrecords import MaskedRecords output = output.view(MaskedRecords) else: output = output.view(np.recarray) return output
def recfromcsv(fname, **kwargs): """
Load ASCII data stored in a comma-separated file.
The returned array is a record array (if ``usemask=False``, see `recarray`) or a masked record array (if ``usemask=True``, see `ma.mrecords.MaskedRecords`).
Parameters ---------- fname, kwargs : For a description of input parameters, see `genfromtxt`.
See Also -------- numpy.genfromtxt : generic function to load ASCII data.
Notes ----- By default, `dtype` is None, which means that the data-type of the output array will be determined from the data.
"""
# Set default kwargs for genfromtxt as relevant to csv import. kwargs.setdefault("case_sensitive", "lower") kwargs.setdefault("names", True) kwargs.setdefault("delimiter", ",") kwargs.setdefault("dtype", None) output = genfromtxt(fname, **kwargs)
usemask = kwargs.get("usemask", False) if usemask: from numpy.ma.mrecords import MaskedRecords output = output.view(MaskedRecords) else: output = output.view(np.recarray) return output
|
