okntool 8.4.2

Command line program to carry out okn related tasks.

OKN TOOL PYTHON PACKAGE LIBRARY MANUAL

Description

This program will draw the graph image in the ABI eye research group web experience recording folder according to config file information and plot type input. Moreover, it can also produce va table csv and html files for both individual and summary data folder.

There are 8 types of okn tool which are:

1. Trial plot which represents sensor timestamp vs pupil displacement/movement graph for each trial.
2. Summary plot which is the combination of all trial plots with the same x and y axis scales.
3. Staircase/progress plot which is the graph image to visualize whether there is okn or not and the visual acuity etdrs calculation steps.
4. Tidy plot, the special plot for all trial protocol in which borders and boundaries are drawn in tidest format.
5. Simpler plot, the special plot for face detection and eye tracker related OKNs.
6. indi_va_table which will create indi_va_table.csv and indi_va_table.html depand on whether there is okn or not in each trials.
7. sum_va_table which will create sum_va_table.csv and sum_va_table.html by using indi_va_table.csv of each folder which is produced by indi_va_table okn tool type.
8. pnm_video_splitter which will split pnm_eye_video.mp4 into pnm_left_eye_video and pnm_right_eye_video.
9. overlay_pupil_detection which will overlay pupil detection on eye videos of opm and plm manager recording.

Installation requirements and guide

Anaconda

To install this program, Anaconda python distributing program and Anaconda Powershell Prompt are needed.
If you do not have Anaconda, please use the following links to download and install:
Download link: https://www.anaconda.com/products/distribution
Installation guide link: https://docs.anaconda.com/anaconda/install/

PIP install

To install okntool, you have to use Anaconda Powershell Prompt.
After that, you can use the okntool from any command prompt.
In Anaconda Powershell Prompt:

pip install okntool

Usage guide

Example usage

okntool -t "(plot_type)" -d "(directory to the folder)" -c "(directory to config file)" -r "(referenced csv)"

plot_type: "trial", "summary" or "staircase"
If you are using "trial" plot type, please -d "(directory to trial folder)".
If you are using "summary" or "staircase" plot type, please -d "(directory to pim_recorded_folder/trials)".
-r is only for "simpler" plot type.

There is a example folder under development folder.
If you want to test this program, you can clone this repository, install okntool and run the following command:
For trial plot

okntool -t "trial" -d "development/example/trials/trials/trial-1-1_disk-condition-1-1"

For summary plot

okntool -t "summary" -d "development/example/trials/trials"

For staircase/progress plot

okntool -t "staircase" -d "development/example/trials/trials"

or

okntool -t "progress" -d "development/example/trials/trials"

For tidy plot

okntool -t "tidy" -d "development/example/trials/trials"

For simpler plot

okntool -t "simpler" -d "(folders which clip folders)"

or

okntool -t "simpler" -d "(folders which clip folders)" -c (simpler plot configuration) -r (referenced csv)

The -c and -r arguments are optional. The okntool has its own built-in config and default referenced csv location which is two folder back of input folder directory. i.e. If the input folder diectory is "(example/result/okn)", the default referenced csv location is "(example/protocol.simpler.csv)".

For indi_va_table

okntool -t "indi_va_table" -d "(folder which contains clip folders each contains decider.json)"

or

okntool -t "indi_va_table" -d "(folder which contains clip folders each contains decider.json)" -p "(template html file)" -n "(decider file name)" -r "(referenced csv)" -o "(output location)"

The -p, -n, -r and -o arguments are optional. The okntool has its own built-in template html, default decider file name and default referenced csv location which is two folder back of input folder directory. i.e. If the input folder diectory is "(example/result/okn)", the default referenced csv location is "(example/protocol.simpler.csv)".

For sum_va_table

okntool -t "sum_va_table" -d "(folder which contains folders each contains indi_va_table.csv)"

or

okntool -t "sum_va_table" -d "(folder which contains folders each contains indi_va_table.csv)" -p "(template html file)" -o "(output location)"

The -p and -o arguments are optional. The okntool has its own built-in template html.

For pnm_video_splitter

okntool -t "pnm_video_splitter" -d "(location of pnm_eye_video.mp4)"

For split_video

okntool -t "split_video" -d "(recording folder directory)"

(or)

okntool -t "sv" -d "(recording folder directory)"

For overlay_pupil_detection

okntool -t "overlay_pupil_detection" -d "(recording folder directory)"

(or)

okntool -t "opd" -d "(recording folder directory)"

If you wanna use okn_detector_summary.csv outside of the recording folder,

okntool -t "overlay_pupil_detection" -d "(recording folder directory)" -sc "(directory to summary csv)"

(or)

okntool -t "opd" -d "(recording folder directory)" -sc "(directory to summary csv)"

oknserver_graph_plot_config.json

This is a built-in config file which contains the information how trial plot, summary plot and progress plot will be drawn.
Please read more details about them in TRIAL_PLOT_README.md, SUMMARY_PLOT_README.md, PROGRESS_PLOT_README.md and TIDY_PLOT_README.md. If you wanna change the parameters of config file, copy and paste the built-in config file into a different directory. After that modify the config file and run the okntool with optional argument -c (directory to new config file).

okntool -t "(plot_type)" -d "(directory to the folder)" -c "(directory to new config file)"

simpler_plot_config.json

This is a built-in config file which contains the information how simpler plot will be drawn.
Please read more details in SIMLER_PLOT_README.md. If you wanna change the parameters of config file, copy and paste the built-in config file into a different directory.

To upgrade version

In Anaconda Powershell Prompt,

pip install -U okntool

or

pip install --upgrade okntool