stems.utils module¶

class stems.utils.FrozenKeyDict(*args, **kwds)[source]¶

Bases: collections.abc.MutableMapping

A dict that doesn’t allow new keys

copy(self)[source]¶

stems.utils.cached_property(prop)[source]¶: Cache a class property (e.g., that requires a lookup)

stems.utils.concat_lists(l)[source]¶

Concatenate all list-like items in l

Parameters: l (list or tuple) – Sequence (but not a str)
Yields: list – Concatenated list items

See also

toolz.concat(): Similar function, but concatenates all iterables
flatten_lists(): Recursive version of this function

stems.utils.dtype_info(arr)[source]¶

Return integer or float dtype info

Parameters: arr (np.ndarray) – Array
Returns: NumPy type information
Return type: numpy.core.getlimits.finfo or numpy.core.getlimits.iinfo
Raises: TypeError – Raised if arr is not a int or float type

stems.utils.find(location, pattern, regex=False)[source]¶

Return a sorted list of files matching pattern

Parameters

location (str or pathlib.Path) – Directory location to search
pattern (str) – Search pattern for files
regex (bool) – True if pattern is a regular expression

Returns

List of file paths for files found

Return type

stems.utils.find_subclasses(cls_)[source]¶

Find subclasses of an object

Parameters: cls_ (class) – A Python class
Returns: Classes that inherit from cls_
Return type: set[class]

stems.utils.flatten_lists(l)[source]¶

Flatten all list-like items in l

Parameters: l (list or tuple) – Sequence (but not a str)
Yields: list – Flattened list items

stems.utils.import_object(string)[source]¶

Import a Python module/class/function from its string

Parameters: string (str) – A Python object path
Returns: The imported object (class/module/function)
Return type: object

stems.utils.list_like(obj)[source]¶

Return True if obj is list-like

List-like includes lists, tuples, and numpy.ndarrays, but not other sequences like str or Mapping.

Parameters: obj (object) – An object
Returns: True if obj is a list (sequence but not str)
Return type: bool

stems.utils.np_promote_all_types(*dtypes)[source]¶

Return the largest NumPy datatype required to hold all types

Parameters: dtypes (iterable) – NumPy datatypes to promote
Returns: Smallest NumPy datatype required to store all input datatypes
Return type: np.dtype

See also

np.promote_types()

stems.utils.register_multi_singledispatch(func, types)[source]¶

Register multiple types for singledispatch

Parameters

func (callable) – Function
types (tuple) – Multiple types to register

Returns

func – Decorated function

Return type

callable

stems.utils.relative_to(one, two)[source]¶

Return the relative path of a file compared to another

Parameters

one (str or Path) – File to return relative path for
two (str or Path) – File one will be relative to

Returns

Relative path of one

Return type

Path

stems.utils.renamed_upon_completion(destination, tmpdir=None, prefix='', suffix=None)[source]¶

Help save/write to file and move upon completion

Parameters

destination (str or Path) – The final intended location of the file
tmpdir (str, optional) – By default, this function will yield a temporary filename in the same directory as destination, but you may specify another location using tmpdir.
prefix (str, optional) – Characters to prefix the temporary file with
suffix (str, optional) – Characters to add at the end of the temporary filename. By default appends “.tmp.” and the process ID

Yields

str – A temporary filename to use during writing/saving/etc

stems.utils.squeeze(l)[source]¶

Squeeze Sequences with 1 item into a scalar

Parameters: l (Sequence) – List, tuple, etc
Returns: Either original object if len(l) != 1 else l[0]
Return type: object

stems.utils.to_number(string)[source]¶: Convert string to most appropriate number

stems.utils.update_nested(d, other)[source]¶

Update a potentially nested dict with another

Parameters

d (dict) – Dictionary to update
other (dict) – Other dict with replacement values

Returns

Updated dict

Return type