Merge pull request #2287 from AntoinePrv/doc-improvement-b

Documentation improvement Part #B

docs/source/conf.py

@@ -39,3 +39,10 @@ def setup(app):
 html_js_files = [
     'goatcounter.js'
 ]
+
+# Automatically link to numpy doc
+extensions += ['sphinx.ext.intersphinx']
+intersphinx_mapping = {
+    "numpy": ("https://numpy.org/doc/stable/", None),
+    "scipy": ("https://docs.scipy.org/doc/scipy/reference", None),
+}

docs/source/histogram.rst

@@ -28,7 +28,7 @@ Basic usage
     *   ``bins = 10``
     *   ``mode = xt::histogram::automatic``
 
-The behavior, in-, and output of ``histogram`` is similar to that of `numpy.histogram <https://docs.scipy.org/doc/numpy/reference/generated/numpy.histogram.html>`_ with that difference that the bin-edges are obtained by a separate function call:
+The behavior, in-, and output of ``histogram`` is similar to that of :any:`numpy.histogram` with that difference that the bin-edges are obtained by a separate function call:
 
 .. code-block:: cpp
 

docs/source/numpy-differences.rst

@@ -47,7 +47,7 @@ array argument is a 0-D argument:
 Meshgrid
 --------
 
-Numpy's version of meshgrid supports two modes: the 'xy' indexing and the 'ij' indexing.
+:any:`numpy.meshgrid` supports two modes: the 'xy' indexing and the 'ij' indexing.
 
 The following code
 
@@ -101,4 +101,4 @@ The reason is an optimization in xtensor, which is to set the strides to ``0`` i
 Array indices
 -------------
 
-Array indices are in xtensor stored as a ``std::vector`` of array indices, whereby each entry corresponds to the array indices of one item. This results in a slightly different usage of ``xt::ravel_indices`` than of ``np.ravel_multi_index``.
+Array indices are in xtensor stored as a ``std::vector`` of array indices, whereby each entry corresponds to the array indices of one item. This results in a slightly different usage of ``xt::ravel_indices`` than of :any:`numpy.ravel_multi_index`.