PDE_ML

Exploring the use of machine learning in PDE simulations.

Requirements

• Python3 sudo apt install python3
• Numpy pip install numpy
• Pandas pip install pandas
• PyTorch (pytorch.org)

Project structure

csvdata.py functions for loading and processing data into format suitable for training and/or validation

descent.py gradient descent for PDE parameter search done on the neural network model - not working well so far

model_lstm.py LSTM model used for PDE acceleration training

model_rl.py LSTM model used for PDE parameter search with reinforcement learning

model_simple.py MLP model used for PDE acceleration training

ode_fun.py set of ODE equations used to solve the PDE system

pSmad_WT_MT_new.mat experimental data

plotter.py plotting tools

prepdata.py data preprocessing script, pulls data from many MAT files and converts it into training and validation CSV files

profile.py script for profiling NNs used for PDE acceleration

run_rl.py PDE parameter search through reinforcement learning

solver.py script to run PDE simulations

train.py training script for PDE acceleration models

Training

To train a model, run train.py with the path to the data:

python train.py --data=/path/to/data

Usage

usage: train.py [-h] [--data DATA] [--lstm] [--use_k] [-j WORKERS]
                [--epochs EPOCHS] [--start-epoch START_EPOCH] [-b BATCH_SIZE]
                [--lr LR] [--momentum MOMENTUM] [--wd WEIGHT_DECAY]
                [-p PRINT_FREQ] [--resume RESUME] [-e] [--seed SEED]
                [--gpu GPU]

Simulation Data Training

optional arguments:
  -h, --help            show this help message and exit
  --data DATA           path to dataset
  --lstm                use lstm
  -j WORKERS, --workers WORKERS
                        number of data loading workers (default: 4)
  --epochs EPOCHS       number of total epochs to run
  --start-epoch START_EPOCH
                        manual epoch number (useful on restarts)
  -b BATCH_SIZE, --batch-size BATCH_SIZE
                        mini-batch size (default: 256)
  --lr LR, --learning-rate LR
                        initial learning rate
  --momentum MOMENTUM   momentum
  --wd WEIGHT_DECAY, --weight-decay WEIGHT_DECAY
                        weight decay (default: 1e-4)
  -p PRINT_FREQ, --print-freq PRINT_FREQ
                        print frequency (default: 10000)
  --resume RESUME       path to latest checkpoint (default: none)
  -e, --evaluate        evaluate model on validation set
  --seed SEED           seed for initializing training.
  --gpu GPU             GPU id to use.