Solves the linear system :py:math:`Ax=b` and returns the result in ``x``.\n\
This method uses LAPACK's ``dgesv`` generic solver.\n\
\n\
THIS VARIANT DOES NOT PERFORM ANY CHECKS ON THE INPUT MATRICES AND IS,\n\
FASTER THEN THE VARIANT NOT ENDING IN ``_``. Use it when you are sure\n\
your input matrices are well-behaved (contiguous, c-style, memory aligned).\n\
.. warning::\n\
\n\
THIS VARIANT DOES NOT PERFORM ANY CHECKS ON THE INPUT MATRICES AND IS,\n\
FASTER THEN THE VARIANT NOT ENDING IN ``_``. Use it when you are sure\n\
your input matrices sizes match.\n\
\n\
You can use this method in two different formats. The first interface\n\
accepts the matrices ``A`` and ``b`` returning ``x``. The second one\n\
...
...
@@ -121,9 +124,11 @@ Solves the linear system :py:math:`Ax=b` and returns the result in ``x``.\n\
This method uses LAPACK's ``dposv`` solver, assuming ``A`` is a symmetric.\n\
positive definite matrix.\n\
\n\
THIS VARIANT DOES NOT PERFORM ANY CHECKS ON THE INPUT MATRICES AND IS,\n\
FASTER THEN THE VARIANT NOT ENDING IN ``_``. Use it when you are sure\n\
your input matrices are well-behaved (contiguous, c-style, memory aligned).\n\
.. warning::\n\
\n\
THIS VARIANT DOES NOT PERFORM ANY CHECKS ON THE INPUT MATRICES AND IS,\n\
FASTER THEN THE VARIANT NOT ENDING IN ``_``. Use it when you are sure\n\
your input matrices sizes match.\n\
\n\
You can use this method in two different formats. The first interface\n\
accepts the matrices ``A`` and ``b`` returning ``x``. The second one\n\
...
...
@@ -157,9 +162,11 @@ Solves the linear system :py:math:`Ax=b` and returns the result in ``x``.\n\
This method solves the linear system via conjugate gradients and assumes\n\
``A`` is a symmetric positive definite matrix.\n\
\n\
THIS VARIANT DOES NOT PERFORM ANY CHECKS ON THE INPUT MATRICES AND IS,\n\
FASTER THEN THE VARIANT NOT ENDING IN ``_``. Use it when you are sure\n\
your input matrices are well-behaved (contiguous, c-style, memory aligned).\n\
.. warning::\n\
\n\
THIS VARIANT DOES NOT PERFORM ANY CHECKS ON THE INPUT MATRICES AND IS,\n\
FASTER THEN THE VARIANT NOT ENDING IN ``_``. Use it when you are sure\n\
your input matrices sizes match.\n\
\n\
You can use this method in two different formats. The first interface\n\
accepts the matrices ``A`` and ``b`` returning ``x``. The second one\n\
...
...
@@ -168,6 +175,70 @@ solution.\n\
"
);
PyDoc_STRVAR(s_pavx_str,"pavx");
PyDoc_STRVAR(s_pavx_doc,
"pavx(input, output) -> None\n\
pavx(input) -> array\n\
\n\
Applies the Pool-Adjacent-Violators Algorithm to ``input``. The ``input``\n\
and ``output`` arrays should have the same size. This is a simplified\n\
C++ port of the isotonic regression code made available at the `University\n\
of Bern website <http://www.imsv.unibe.ch/content/staff/personalhomepages/duembgen/software/isotonicregression/index_eng.html>`_.\n\
\n\
You can use this method in two different formats. The first interface\n\
accepts the 1D float arrays ``input`` and ``output``. The second one\n\
accepts the input array ``input`` and allocates a new ``output`` array\n\
which is returned. In such a case, the ``output`` is a 1D float array\n\
with the same length as ``input``.\n\
");
PyDoc_STRVAR(s_pavx_nocheck_str,"pavx_");
PyDoc_STRVAR(s_pavx_nocheck_doc,
"pavx(input, output) -> None\n\
\n\
Applies the Pool-Adjacent-Violators Algorithm to ``input`` and places the\n\
result on ``output``. The ``input`` and ``output`` arrays should be 1D\n\
float arrays with the same length.\n\
\n\
This is a simplified C++ port of the isotonic regression code\n\
made available at the `University of Bern website <http://www.imsv.unibe.ch/content/staff/personalhomepages/duembgen/software/isotonicregression/index_eng.html>`_.\n\
\n\
.. warning::\n\
\n\
THIS VARIANT DOES NOT PERFORM ANY CHECKS ON THE INPUT MATRICES AND IS,\n\
FASTER THEN THE VARIANT NOT ENDING IN ``_``. Use it when you are sure\n\
your input and output vector sizes match.\n\
\n\
");
PyDoc_STRVAR(s_pavx_width_str,"pavxWidth");
PyDoc_STRVAR(s_pavx_width_doc,
"pavxWidth(input, output) -> array\n\
\n\
Applies the Pool-Adjacent-Violators Algorithm to ``input`` and places the\n\
result on ``output``. The ``input`` and ``output`` arrays should be 1D\n\
float arrays with the same length.\n\
\n\
The width array (64-bit unsigned integer 1D) is returned and has the\n\
Applies the Pool-Adjacent-Violators Algorithm to ``input`` and sets the\n\
result on ``output``. The ``input`` and ``output`` arrays should be 1D\n\
float arrays of the same length.\n\
\n\
This is a simplified C++ port of the isotonic regression code\n\
made available at the `University of Bern website <http://www.imsv.unibe.ch/content/staff/personalhomepages/duembgen/software/isotonicregression/index_eng.html>`_.\n\
\n\
The width and height arrays are returned. The width array is a 64-bit\n\
**unsigned integer** 1D array, while the height array (second component\n\
of the returned tuple) is a 64-bit **float** 1D array of the same size.\n\
staticconstchar*C_PAVX_DOC="Applies the Pool-Adjacent-Violators Algorithm. The input and output arrays should have the same size. This is a simplified C++ port of the isotonic regression code made available at http://www.imsv.unibe.ch/content/staff/personalhomepages/duembgen/software/isotonicregression/index_eng.html.";
staticconstchar*C_PAVX__DOC="Applies the Pool-Adjacent-Violators Algorithm. The input and output arrays should have the same size. Arguments are not checked! This is a simplified C++ port of the isotonic regression code made available at http://www.imsv.unibe.ch/content/staff/personalhomepages/duembgen/software/isotonicregression/index_eng.html.";
staticconstchar*P_PAVX_DOC="Applies the Pool-Adjacent-Violators Algorithm. The output array is allocated and returned. This is a simplified C++ port of the isotonic regression code made available at http://www.imsv.unibe.ch/content/staff/personalhomepages/duembgen/software/isotonicregression/index_eng.html.";
staticconstchar*P_PAVXWIDTH_DOC="Applies the Pool-Adjacent-Violators Algorithm. The input and output arrays should have the same size. The width array is returned. This is a simplified C++ port of the isotonic regression code made available at http://www.imsv.unibe.ch/content/staff/personalhomepages/duembgen/software/isotonicregression/index_eng.html.";
staticconstchar*P_PAVXWIDTHHEIGHT_DOC="Applies the Pool-Adjacent-Violators Algorithm. The input and output arrays should have the same size. The width and height arrays are returned. This is a simplified C++ port of the isotonic regression code made available at http://www.imsv.unibe.ch/content/staff/personalhomepages/duembgen/software/isotonicregression/index_eng.html.";
