Anurag14 / Inflow-Prediction-Bhakra

Code for papers Optimal Reservoir Operations using Long Short-Term Memory Network and Bi-directional Storage Capacity and Elevation Level Calculator for Reservoir Operation Management

Inflow-Prediction-Bhakra

This repository contains codes and data for following publications:

Optimal Reservior Operation using Long Short-Term Memory network

Bi-directional Storage Capacity and Elevation Level Calculator for Reservoir Operation Management

System Requirements

python 3.6+
tensorflow-gpu
keras
pandas
numpy
matplotlib
sklearn

How to run

Install anaconda from this on your pc and run following commands

conda create --name tf-gpu
conda activate tf-gpu
conda install -c aaronzs tensorflow-gpu
conda install -c anaconda cudatoolkit
conda install -c anaconda cudnn
conda install keras-gpu
pip install pandas
pip install sklearn
pip install matplotlib

Then run to download the git repository

git clone https://github.com/Anurag14/Inflow-Prediction-Bhakra
cd Inflow-Prediction-Bhakra

To auto generate model graphs

download graphviz2.38 from here then add its executable to $PATH variable.

import os
os.environ["PATH"] += os.pathsep + 'C:/Program Files (x86)/Graphviz2.38/bin/'

or run following if above doesn't seem to work

pip install pydot_ng
conda install graphviz

TL;DR install graphviz via conda and add its path to $PATH instead of building it from base.

Update: As of writing this library there seems to be a open issue here

Same quick fix is applied by hacking into the keras library ../keras/utils/vis_utils.py and change

 import pydot

to

import pydot_ng as pydot

Then run LSTM/predict.py

python LSTM/predict.py

Citations: If you use this code in your work. Please cite the papers this and this

Enjoy!