[Ifeffit] How to use Larch as backend for Athena and Artemis on Mac (Re: Ifeffit Digest, Vol 174, Issue 8)

[Luke Higgins]

Dear All,

Following the excellent guide that Hiroyuki & Bruce gave for running larch on Demeter on Windows (Ifeffit Digest, Vol 174, Issue 8). The following steps worked for me to run larch/Demeter on MacOS (Sierra  10.12.6).

1) Install Demeter using the advice from: https://bruceravel.github.io/demeter/documents/SinglePage/macinstallation.html
2) Installed Anaconda and larch using the following guide: http://cars.uchicago.edu/xraylarch/installation/index.html (Python & larch from source is fine: but then use Hiroyuki’s step 3&4 advice [attached below]) 

3) In a new terminal window check larch by: 
       larch --version     
       You should see "larch command-line version 0.2".
       
4) Running larch and Demeter, run in a new terminal window: 
       larch --port=4966 –rc
       
5) Open Demeter.pm in a text editor (which for me was buried in: /opt/local/lib/perl5/site_perl/5.24/darwin-thread-multi-2level/Demeter.pm) and change line 44 to: 

   $ENV{DEMETER_BACKEND} ||= 'larch';
   
6) This has changed both Athena and Artemis to using larch – haven’t fully tested yet, but looks good.

Best Wishes,

Luke

(from Hhiroyuki, installing python and larch from source files:)
3. Install unofficial Numpy and Scipy, Python libraries
3-1. Download numpy‑1.13.1+mkl‑cp36‑cp36m‑win_amd64.whl and scipy‑0.19.1‑cp36‑cp36m‑win_amd64.whl via http://www.lfd.uci.edu/~gohlke/pythonlibs/
3-2. pip install numpy-1.13.1+mkl-cp36-cp36m-win_amd64.whl
3-3. pip install scipy‑0.19.1‑cp36‑cp36m‑win_amd64.whl

4. Install Larch and its dependencies
4-1. Start a command prompt
4-2. git clone https://github.com/xraypy/xraylarch.git
4-3. cd xraylarch
4-4. pip install lmfit
4-5. pip install h5py
4-6. pip install pillow
4-7. pip install requests
4-8. pip install sqlalchemy
4-9. pip install matplotlib
4-10. pip install pypiwin32
4-11. pip install wxpython
4-12. pip install wxmplot
4-13. pip install wxutils
4-14. python setup.py install
4-15. larch --version
4-16. You will see "larch command-line version 0.2".
Note: I installed Larch dependencies except for pyFAI, CifFile, fabio, tomopy, scikit-image.