# README

## Installation
Tested on Windows 10 (1909); Anaconda version 4.8.1
Install Miniconda from: https://docs.conda.io/en/latest/miniconda.html
Time <5min

In a fresh conda environment, install the dependecies using:
```powershell
conda create -n test
conda activate test
conda install python=3.8.5 ##3.8+ is a hard requirment; pickling protocol is not backwards compatible
pip install git+https://github.com/jselvan/simianpy.git@0.1.4
```
Time <1 min

simianpy will install all necessary dependencies at the most recent version.  Tested versions for the major dependencies used here are: pandas=1.1.0 matplotlib=3.1.1 numpy=1.19.1 tables=3.6.1

## The code
Navigate to https://gin.g-node.org/everling_lab_marmosets/marmoset_face_processing
Download Data.h5, get_fixations_and_saccades.py and plots.py to the same directory
Activate the environment and execute the scripts
```powershell
conda activate test
python get_fixations_and_saccades.py
python plots.py
```

### get_fixations_and_saccades.py
Here, fixations and saccades are detected block by block using DetectFixations and DetectSaccades from simianpy.  The criteria used are outlined in the manuscript.
Linear convolution of an 11 sample Hamming window is used to smooth the eye data before fixation detection.
Extracted fixation and saccade metrics are stored back into Data.h5

### plots.py
The data extracted in the above step is plotted using matplotlib into 2 png files, that are saved in the same directory