After updating my RP3 to kernel release/version 4.14.52-v7+/#1123, trying to install or update packages such as pandas, numpy, etc. with conda (arm71, v3.16.0) causes an error:
Error: Could not find some dependencies for <package>: blas * openblas
Installing packages with pip still works fine.
The best option for you will be to use berryconda, as standard conda channels lack the support of ARM-based builds
There are 2 approach to my understanding. I'm not 100% that it resolves your issue, so any feedback is of interest.
(Some info were found here: Scientific Python for Raspberry Pi)
I have found that some packages are available using apt-get :
sudo apt-get install python-numpy python-scipy python-pandas python-h5py
I realize that it is not strictly conda usage but it can do the trick.
Otherwise, you can use the follwing commands (here exmaple is given with scipy). Please be aware that you might face compatibility issues between packages or python versions. Some packages might not be available for ARM
anaconda search -t conda scipy
This gives you something like:
Using binstar api site https://api.anaconda.org
Run 'anaconda show <USER/PACKAGE>' to get more details:
Packages:
Name | Version | Package Types | Platforms
------------------------- | ------ | --------------- | ---------------
Alges/scipy | 0.19.1 | conda | linux-64, win-64
: Scientific Library for Python
BioBuilds/scipy | 1.0.0 | conda | linux-ppc64le
... some lines removed ... here is a result with a grep on 'arm'
RaspberryPi/scipy | 0.13.2 | conda | linux-armv6l
alorenzo175/scipy | 0.18.0 | conda | linux-armv7l
compass/scipy | 0.19.0 | conda | linux-armv7l
poppy-project/scipy | 0.17.0 | conda | linux-armv7l
rpi/scipy | 1.0.0 | conda | linux-armv6l, linux-armv7l
then you can do this:
anaconda show rpi/scipy
which gives:
Using binstar api site https://api.anaconda.org
Name: scipy
Summary: Scientific Library for Python
Access: public
Package Types: conda
Versions:
+ 0.18.1
+ 0.19.0
+ 0.19.1
+ 1.0.0
To install this package with conda run:
conda install --channel https://conda.anaconda.org/rpi scipy
Then all you have to do is execute the said command:
conda install --channel https://conda.anaconda.org/rpi scipy
Fetching package metadata: ......
Solving package specifications: .
Error: Unsatisfiable package specifications.
Generating hint:
[ COMPLETE ]|#################################################################################################################################################################################################################| 100%
Hint: the following packages conflict with each other:
- scipy
- python 3.4*
Use 'conda info scipy' etc. to see the dependencies for each package.
Printing some infos:
conda info scipy
Fetching package metadata: ....
scipy 0.16.0 np19py34_1
-----------------------
file name : scipy-0.16.0-np19py34_1.tar.bz2
name : scipy
version : 0.16.0
build number: 1
build string: np19py34_1
channel : defaults
size : 24.4 MB
date : 2015-08-20
license : BSD
md5 : a1f8618171817de092b818a3739ef95c
installed environments:
dependencies:
blas * openblas
numpy 1.9*
openblas 0.2.14
python 3.4*
scipy 0.16.0 np19py27_1
-----------------------
file name : scipy-0.16.0-np19py27_1.tar.bz2
name : scipy
version : 0.16.0
build number: 1
build string: np19py27_1
channel : defaults
size : 25.1 MB
date : 2015-08-20
license : BSD
md5 : bc3882c6770d1dc7e3e5495266002d56
installed environments:
dependencies:
blas * openblas
numpy 1.9*
openblas 0.2.14
python 2.7*
