Version 1.5.0
git clone https://github.com/badw/arcs.git
cd arcs
pip install .
an example Jupyter Notebook can be found here:
./examples/example.ipynb
an example DFT data is found here:
./app/data/dft_data.json
which was run using VASP and the SCAN meta-GGA functional.
simple usage:
from arcs.generate import GraphGenerator
from arcs.traversal import Traversal
from arcs.generate import GenerateInitialConcentrations
graph = GraphGenerator().from_file(
filename='../app/data/dft_data.json',
temperature=248,
pressure=20,
max_reaction_length=5
)
concentrations = GenerateInitialConcentrations(graph=graph).update_ic(
{'H2O':30,'O2':10,'SO2':10,'H2S':10,'NO2':10}
)
t = Traversal(graph=graph)
results = t.sample(
initial_concentrations=concentrations,
ncpus=4,
nsamples=1000
)
results can then be analysed with arcs.analysis.AnalyseSampling
- reaction statistics
from arcs.analysis import AnalyseSampling
import pandas as pd
analysis = AnalyseSampling()
stats = pd.Series(analysis.reaction_statistics(data)).sort_values(ascending=False)
stats.head(10)
1 H2O + 1 SO2 = 1 H2SO3 270 2 H2 + 1 O2 = 2 H2O 227 3 O2 + 2 H2S = 2 H2O + 2 SO2 163 . . .
- Mean average data
average_data = pd.DataFrame(analysis.average_sampling(data))
average_data = average_data.loc[~(average_data==0).all(axis=1)]
average_data.sort_values(by='diff',inplace=True)
average_data.round(2)
compound initial mean diff sem std var H2S 10.0 4.88 -5.12 0.10 4.75 22.53 NO2 10.0 6.19 -3.81 0.10 4.85 23.48 O2 10.0 6.24 -3.76 0.12 5.76 33.18 . . .
The app is created using the plotly DASH (https://github.com/plotly/dash) framework.
The arcs-app can be run from the app directory in the terminal through;
cd app
python arcs-app.py
This project was funded by and carried out with collaboration from Equinor, Total Energies, and Shell
A forked and modified version of ARCS is available at https://github.com/equinor/arcs
This version of ARCS is an "under-development" version of ARCS.