diff options
| author | Daniel Ching <carterbox@users.noreply.github.com> | 2019-11-15 12:00:23 -0600 |
|---|---|---|
| committer | Matti Picus <matti.picus@gmail.com> | 2019-11-15 11:00:23 -0700 |
| commit | b2b45fe9b65db490b84e90fea587c5bb867906eb (patch) | |
| tree | 77168efe7adf9239fa7e3c7af74ac1511a08a529 /numpy | |
| parent | e61f5b0e8a981fd4623dc18fbaa5caec1323882f (diff) | |
| download | numpy-b2b45fe9b65db490b84e90fea587c5bb867906eb.tar.gz | |
DOC: Note FFT type promotion (#14912)
* DOC: Note FFT type promotion
The NumPy FFT implementation will promote `float32` types to `float64` types. This may not be desired for some applications, but the SciPy implementation supports more data types.
Closes #14892
Diffstat (limited to 'numpy')
| -rw-r--r-- | numpy/fft/__init__.py | 8 |
1 files changed, 8 insertions, 0 deletions
diff --git a/numpy/fft/__init__.py b/numpy/fft/__init__.py index fe95d8b17..37b3f0da6 100644 --- a/numpy/fft/__init__.py +++ b/numpy/fft/__init__.py @@ -118,8 +118,16 @@ The inverse DFT is defined as It differs from the forward transform by the sign of the exponential argument and the default normalization by :math:`1/n`. +Type Promotion +-------------- + +`numpy.fft` promotes ``float32`` and ``complex64`` arrays to ``float64`` and +``complex128`` arrays respectively. For an FFT implementation that does not +promote input arrays, see `scipy.fftpack`. + Normalization ------------- + The default normalization has the direct transforms unscaled and the inverse transforms are scaled by :math:`1/n`. It is possible to obtain unitary transforms by setting the keyword argument ``norm`` to ``"ortho"`` (default is |