diff options
Diffstat (limited to 'numpy/lib/function_base.py')
| -rw-r--r-- | numpy/lib/function_base.py | 83 |
1 files changed, 56 insertions, 27 deletions
diff --git a/numpy/lib/function_base.py b/numpy/lib/function_base.py index c7ddbdb8d..af5a6e45c 100644 --- a/numpy/lib/function_base.py +++ b/numpy/lib/function_base.py @@ -8,7 +8,7 @@ import numpy as np import numpy.core.numeric as _nx from numpy.core import transpose from numpy.core.numeric import ( - ones, zeros, arange, concatenate, array, asarray, asanyarray, empty, + ones, zeros_like, arange, concatenate, array, asarray, asanyarray, empty, ndarray, around, floor, ceil, take, dot, where, intp, integer, isscalar, absolute ) @@ -593,7 +593,7 @@ def piecewise(x, condlist, funclist, *args, **kw): not isinstance(condlist[0], (list, ndarray)) and x.ndim != 0): condlist = [condlist] - condlist = array(condlist, dtype=bool) + condlist = asarray(condlist, dtype=bool) n = len(condlist) if n == n2 - 1: # compute the "otherwise" condition. @@ -606,7 +606,7 @@ def piecewise(x, condlist, funclist, *args, **kw): .format(n, n, n+1) ) - y = zeros(x.shape, x.dtype) + y = zeros_like(x) for cond, func in zip(condlist, funclist): if not isinstance(func, collections.abc.Callable): y[cond] = func @@ -671,11 +671,22 @@ def select(condlist, choicelist, default=0): raise ValueError("select with an empty condition list is not possible") choicelist = [np.asarray(choice) for choice in choicelist] - choicelist.append(np.asarray(default)) + + try: + intermediate_dtype = np.result_type(*choicelist) + except TypeError as e: + msg = f'Choicelist elements do not have a common dtype: {e}' + raise TypeError(msg) from None + default_array = np.asarray(default) + choicelist.append(default_array) # need to get the result type before broadcasting for correct scalar # behaviour - dtype = np.result_type(*choicelist) + try: + dtype = np.result_type(intermediate_dtype, default_array) + except TypeError as e: + msg = f'Choicelists and default value do not have a common dtype: {e}' + raise TypeError(msg) from None # Convert conditions to arrays and broadcast conditions and choices # as the shape is needed for the result. Doing it separately optimizes @@ -846,7 +857,7 @@ def gradient(f, *varargs, axis=None, edge_order=1): Returns ------- gradient : ndarray or list of ndarray - A set of ndarrays (or a single ndarray if there is only one dimension) + A list of ndarrays (or a single ndarray if there is only one dimension) corresponding to the derivatives of f with respect to each dimension. Each derivative has the same shape as f. @@ -1290,7 +1301,7 @@ def _interp_dispatcher(x, xp, fp, left=None, right=None, period=None): @array_function_dispatch(_interp_dispatcher) def interp(x, xp, fp, left=None, right=None, period=None): """ - One-dimensional linear interpolation. + One-dimensional linear interpolation for monotonically increasing sample points. Returns the one-dimensional piecewise linear interpolant to a function with given discrete data points (`xp`, `fp`), evaluated at `x`. @@ -1337,8 +1348,8 @@ def interp(x, xp, fp, left=None, right=None, period=None): -------- scipy.interpolate - Notes - ----- + Warnings + -------- The x-coordinate sequence is expected to be increasing, but this is not explicitly enforced. However, if the sequence `xp` is non-increasing, interpolation results are meaningless. @@ -2191,15 +2202,14 @@ class vectorize: ufunc, otypes = self._get_ufunc_and_otypes(func=func, args=args) # Convert args to object arrays first - inputs = [array(a, copy=False, subok=True, dtype=object) - for a in args] + inputs = [asanyarray(a, dtype=object) for a in args] outputs = ufunc(*inputs) if ufunc.nout == 1: - res = array(outputs, copy=False, subok=True, dtype=otypes[0]) + res = asanyarray(outputs, dtype=otypes[0]) else: - res = tuple([array(x, copy=False, subok=True, dtype=t) + res = tuple([asanyarray(x, dtype=t) for x, t in zip(outputs, otypes)]) return res @@ -2268,13 +2278,13 @@ class vectorize: def _cov_dispatcher(m, y=None, rowvar=None, bias=None, ddof=None, - fweights=None, aweights=None): + fweights=None, aweights=None, *, dtype=None): return (m, y, fweights, aweights) @array_function_dispatch(_cov_dispatcher) def cov(m, y=None, rowvar=True, bias=False, ddof=None, fweights=None, - aweights=None): + aweights=None, *, dtype=None): """ Estimate a covariance matrix, given data and weights. @@ -2325,6 +2335,11 @@ def cov(m, y=None, rowvar=True, bias=False, ddof=None, fweights=None, weights can be used to assign probabilities to observation vectors. .. versionadded:: 1.10 + dtype : data-type, optional + Data-type of the result. By default, the return data-type will have + at least `numpy.float64` precision. + + .. versionadded:: 1.20 Returns ------- @@ -2400,13 +2415,16 @@ def cov(m, y=None, rowvar=True, bias=False, ddof=None, fweights=None, if m.ndim > 2: raise ValueError("m has more than 2 dimensions") - if y is None: - dtype = np.result_type(m, np.float64) - else: + if y is not None: y = np.asarray(y) if y.ndim > 2: raise ValueError("y has more than 2 dimensions") - dtype = np.result_type(m, y, np.float64) + + if dtype is None: + if y is None: + dtype = np.result_type(m, np.float64) + else: + dtype = np.result_type(m, y, np.float64) X = array(m, ndmin=2, dtype=dtype) if not rowvar and X.shape[0] != 1: @@ -2486,12 +2504,14 @@ def cov(m, y=None, rowvar=True, bias=False, ddof=None, fweights=None, return c.squeeze() -def _corrcoef_dispatcher(x, y=None, rowvar=None, bias=None, ddof=None): +def _corrcoef_dispatcher(x, y=None, rowvar=None, bias=None, ddof=None, *, + dtype=None): return (x, y) @array_function_dispatch(_corrcoef_dispatcher) -def corrcoef(x, y=None, rowvar=True, bias=np._NoValue, ddof=np._NoValue): +def corrcoef(x, y=None, rowvar=True, bias=np._NoValue, ddof=np._NoValue, *, + dtype=None): """ Return Pearson product-moment correlation coefficients. @@ -2525,6 +2545,11 @@ def corrcoef(x, y=None, rowvar=True, bias=np._NoValue, ddof=np._NoValue): Has no effect, do not use. .. deprecated:: 1.10.0 + dtype : data-type, optional + Data-type of the result. By default, the return data-type will have + at least `numpy.float64` precision. + + .. versionadded:: 1.20 Returns ------- @@ -2616,7 +2641,7 @@ def corrcoef(x, y=None, rowvar=True, bias=np._NoValue, ddof=np._NoValue): # 2015-03-15, 1.10 warnings.warn('bias and ddof have no effect and are deprecated', DeprecationWarning, stacklevel=3) - c = cov(x, y, rowvar) + c = cov(x, y, rowvar, dtype=dtype) try: d = diag(c) except ValueError: @@ -4085,9 +4110,12 @@ def trapz(y, x=None, dx=1.0, axis=-1): Returns ------- - trapz : float - Definite integral as approximated by trapezoidal rule. - + trapz : float or ndarray + Definite integral of 'y' = n-dimensional array as approximated along + a single axis by the trapezoidal rule. If 'y' is a 1-dimensional array, + then the result is a float. If 'n' is greater than 1, then the result + is an 'n-1' dimensional array. + See Also -------- sum, cumsum @@ -4230,7 +4258,7 @@ def meshgrid(*xi, copy=True, sparse=False, indexing='xy'): See Also -------- mgrid : Construct a multi-dimensional "meshgrid" using indexing notation. - ogrid : Construct an open multi-dimensional "meshgrid" using indexing + ogrid : Construct an open multi-dimensional "meshgrid" using indexing notation. Examples @@ -4259,7 +4287,8 @@ def meshgrid(*xi, copy=True, sparse=False, indexing='xy'): >>> y = np.arange(-5, 5, 0.1) >>> xx, yy = np.meshgrid(x, y, sparse=True) >>> z = np.sin(xx**2 + yy**2) / (xx**2 + yy**2) - >>> h = plt.contourf(x,y,z) + >>> h = plt.contourf(x, y, z) + >>> plt.axis('scaled') >>> plt.show() """ |