@@ -28,22 +28,21 @@ or ``conda``::
28
28
29
29
conda install scikit-learn
30
30
31
- ** We don't recommend installing scipy or numpy using pip on Linux **,
32
- as this will involve a lengthy build process with many dependencies.
33
- Without careful configuration, building numpy yourself is likely to lead to an
34
- installation that is much slower than it should be.
35
- If you are using Linux, consider using your package manager to install
36
- scikit-learn. It is usually the easiest way, but might not provide the newest
37
- version.
38
- If you haven't already installed numpy and scipy and can't install them via
39
- your operating system, we recommended using a third-party distribution.
31
+ If you have not installed NumPy or SciPy yet, you can also install these using
32
+ conda or pip. When using pip, please ensure that * binary wheels * are used,
33
+ and NumPy and SciPy are not recompiled from source, which can happen when using
34
+ particular configurations of operating system and hardware (such as Linux on
35
+ a Raspberry Pi).
36
+ Building numpy and scipy from source can be complex (especially on Windows) and
37
+ requires careful configuration to ensure that they link against an optimized
38
+ implementation of linear algebra routines.
39
+ Instead, use a third-party distribution as described below .
40
40
41
41
If you must install scikit-learn and its dependencies with pip, you can install
42
- it as ``scikit-learn[alldeps] ``. We strongly recommend against doing this
43
- unless you are familiar with how to correctly build numpy and scipy. The most
44
- common use case for this is in a ``requirements.txt `` file used as part of an
45
- automated build process for a PaaS application or a Docker image. This option
46
- is not intended for manual installation from the command line.
42
+ it as ``scikit-learn[alldeps] ``. The most common use case for this is in a
43
+ ``requirements.txt `` file used as part of an automated build process for a PaaS
44
+ application or a Docker image. This option is not intended for manual
45
+ installation from the command line.
47
46
48
47
Third-party Distributions
49
48
==========================
0 commit comments