Skip to content

Python - MATLAB/Octave

Running python inside MATLAB

[The following needs to be edited, example needs to be added, and it needs to be verified again. ]

Using MATLAB to run rs_py

params = py.dict(pyargs( ...
    "input_path", "/path/to/input", ...
    "output_dir", "/path/to/output", ...
    "comparison_type", "triadic" ...
));

meta = py.dict(pyargs( ...
    "subject", "S4", ...
    "exp_name", "demo_experiment", ...
    "num_trials", double(1110), ...
    "num_sessions", double(10) ...
));

params{"metadata"} = meta;

mod = py.importlib.import_module("src.rs_py.scripts.write_choice_file_detailed");
mod.run(params);

json_str = '{"input_path":"/path/to/input","output_dir":"/path/to/output","comparison_type":"triadic","metadata":{"subject":"S4","exp_name":"demo_experiment","num_trials":1110,"num_sessions":10}}';

mod = py.importlib.import_module("src.rs_py.scripts.write_choice_file_detailed");
mod.run(json_str);

Dependencies for MATLAB environment

Create a conda environment with Python 3.10 or higher. 

conda create -n rs_env python=3.10

Install required packages into the environment via pip 

pip install numpy scipy pandas

Find the path to your virtual environment 

conda env list
You may see output resembling the following: 
# conda environments:
#
# * -> active
# + -> frozen
                         /Users/suniyya/fsl
base                     /Users/suniyya/miniconda3
matlab_env           *   /Users/suniyya/miniconda3/envs/rs_env
mkdocs_env               /Users/suniyya/miniconda3/envs/mkdocs_env

Download rs-software from Github. Make sure that src in added to their path.

Open the MATLAB console and set the environment as follows: from inside rs-software

Set python env 

pyenv(Version='/Users/suniyya/miniconda3/envs/rs_env');

import module. 3 modules to choose.

Add a field to coords file to inlcude used params eg iterations etc.

JSON example first time around. But otherwise pydict.

TEST to see if missingh options fail from None. in MATLAB.

Running MATLAB inside python