From 710578fd9b8ec902313ace9ee7adca3e205ecef9 Mon Sep 17 00:00:00 2001 From: Artyom Nikitin Date: Mon, 7 Jun 2021 23:09:21 +0300 Subject: [PATCH] doc: update readme --- README.md | 24 +++++++++++++++++++++++- 1 file changed, 23 insertions(+), 1 deletion(-) diff --git a/README.md b/README.md index ff328cb..38445f8 100644 --- a/README.md +++ b/README.md @@ -1,6 +1,6 @@ # Probabilistic Substance Quality Index (PSQI) -This repository contains experimental results for the paper that introduced PSQI index and applied it to water quality measurements conducted in New Moscow region, Russia. +This repository contains experimental results and source code for the paper that introduced PSQI index and applied it to water quality measurements conducted in New Moscow region, Russia. # Results @@ -20,3 +20,25 @@ This interactive map of measurements can be opened with [Kepler.gl](https://kepl 3. Select "Load Map using URL"; 4. Paste [this URL](https://raw.githubusercontent.com/tzoiker/psqi/master/results/map.json) of the raw map.json content file and click "Fetch". Alternatively, one may download map.json file and upload it directly after the step 2. + +# Source code + +* `src/data` - raw data used for experiments; +* `src/psqi` - main code used for experiments; +* `src/results` - output of the code execution; +* `src/playground.ipynb` - jupyter notebook for interactive execution of the code. + +## Preparation + +1. Install [Conda](https://conda.io/projects/conda/en/latest/user-guide/install/index.html). +2. Create python environment with `conda create -n psqi python=3.6`. +3. Install gpytorch with `conda install --name psqi gpytorch==0.3.6 -c gpytorch`. +4. (Optional) To enable GPU support, install corresponding [pytoch](https://pytorch.org/get-started/previous-versions/#v180) version. +5. Activate environment with `conda activate psqi`. +6. Install other dependencies with `pip install -r requirements.txt`. +7. Install jupyter notebook kernel with `python -m ipykernel install --name=psqi`. + +## Running the code + +Run jupyter notebook with `jupyter notebook` and open `src/playground.ipynb` in the automatically opened browser window. +Now jupyter cells can be sequentially executed to reproduce the results.