---
license: cc-by-nc-4.0
language:
- en
tags:
- multiphase flows
size_categories:
- 10K<n<100K
---
**Overview**
Scientific machine learning (SciML) is a promising strategy for designing multiphase flow solvers, but it requires a large dataset. This paper presents a comprehensive dataset created from 11,000 simulations, both in 2D and 3D, using the Lattice Boltzmann method~(LBM). The dataset captures intricate physics by varying factors such as surface tension, density, and viscosity of fluids. These simulations, comprising 1 million time snapshots, provide extensive data on two-fluid behavior. By making this dataset publicly available, we aim to encourage SciML research and its applications in complex fluid systems, facilitating the creation of more precise and efficient SciML frameworks for multiphysics applications. Our dataset spans multiple orders of magnitude of Reynolds and Bond numbers, density, and viscosity ratios, making it exceptionally rich and valuable for understanding multiphase flows.


**Dataset Information**

Our library contains simulated data from both 2d and 3d simulaions. The library has four categories (with corresponding sample sizes):

```2d bubble - 5000```
```2d drop - 5000```
```3d bubble - 500```
```3d drop - 500```

Due to size of the datasets and to facilitate easier downloads, we have divided the simulations under each category into groups. Further, we have created tar files (for 2d) and splitted tar files (for 3d) to reduce the download sizes. The following steps show how to extract these into ```*.npz.tar.gz``` files.

In addition to the four categories listed above, we have included a new dataset that was used in the MPFBench paper, located at `2Dbubble/mpf_paper_dataset`. This dataset includes text and train splitter data for S1 through S6 input-output time mappings.


***2d***
```
tar -xzvf dropGroup1_npz_*.tar.gz
tar -xzvf dropGroup2_npz_*.tar.gz
tar -xzvf dropGroup3_npz_*.tar.gz
tar -xzvf dropGroup4_npz_*.tar.gz
tar -xzvf dropGroup5_npz_*.tar.gz

tar -xzvf bubbleGroup1_npz_*.tar.gz
tar -xzvf bubbleGroup2_npz_*.tar.gz
tar -xzvf bubbleGroup3_npz_*.tar.gz
tar -xzvf bubbleGroup4_npz_*.tar.gz
tar -xzvf bubbleGroup5_npz_*.tar.gz
```


***3d***
```
Step - 1 : Merge split tarred files into tar.gz files

    cat 3DbubbleGroup1_npz_* > 3DbubbleGroup1_npz.tar.gz
    cat 3DbubbleGroup2_npz_* > 3DbubbleGroup2_npz.tar.gz
    cat 3DbubbleGroup3_npz_* > 3DbubbleGroup3_npz.tar.gz
    cat 3DbubbleGroup4_npz_* > 3DbubbleGroup4_npz.tar.gz
    cat 3DbubbleGroup5_npz_* > 3DbubbleGroup5_npz.tar.gz

    cat 3DdropGroup1_npz_* > 3DdropGroup1_npz.tar.gz
    cat 3DdropGroup2_npz_* > 3DdropGroup2_npz.tar.gz
    cat 3DdropGroup3_npz_* > 3DdropGroup3_npz.tar.gz
    cat 3DdropGroup4_npz_* > 3DdropGroup4_npz.tar.gz
    cat 3DdropGroup5_npz_* > 3DdropGroup5_npz.tar.gz

Step 2 : Create the *npz.tar.gz files

    tar -xzvf 3DdropGroup1_npz.tar.gz
    tar -xzvf 3DdropGroup2_npz.tar.gz
    tar -xzvf 3DdropGroup3_npz.tar.gz
    tar -xzvf 3DdropGroup4_npz.tar.gz
    tar -xzvf 3DdropGroup5_npz.tar.gz

    tar -xzvf 3DbubbleGroup1_npz.tar.gz
    tar -xzvf 3DbubbleGroup2_npz.tar.gz
    tar -xzvf 3DbubbleGroup3_npz.tar.gz
    tar -xzvf 3DbubbleGroup4_npz.tar.gz
    tar -xzvf 3DbubbleGroup5_npz.tar.gz

```

Refer to the directory structure section below to understand how these files have been staged inside huggingface.

**Data Preparation**

Please refer to our [github](https://github.com/baskargroup/mpfbench-tools/tree/master/DataPrep) page that provides codes for doing data preparation for SciML models:

```
https://github.com/baskargroup/mpfbench-tools/tree/master/DataPrep
```

**License**
```
CC-BY-NC-4.0
```

**How to download files from Huggingface**

To run the example code, you need to install the following package:

```bash
pip install huggingface_hub
```

The following script demonstrates how to download a directory from the Hugging Face Hub:

```python 
from huggingface_hub import HfApi, hf_hub_download
import os
import shutil

REPO_ID = "BGLab/mpf-bench"
DIRECTORY = "2Dbubble/bubbleGroup1_npz"

# Initialize the Hugging Face API
api = HfApi()

# List files in the directory
files_list = api.list_repo_files(repo_id=REPO_ID, repo_type="dataset")

# Filter the files in the specified directory
files_to_download = [f for f in files_list if f.startswith(DIRECTORY)]

# Create local directory if it doesn't exist
os.makedirs(DIRECTORY, exist_ok=True)

# Download each file
for file in files_to_download:
    file_path = hf_hub_download(repo_id=REPO_ID, filename=file, repo_type="dataset")
    # Copy the file to the local directory using shutil.copy2
    shutil.copy2(file_path, os.path.join(DIRECTORY, os.path.basename(file_path)))

print("Files downloaded successfully.")

```

**Directory Structure**
```
main/
├── 2Dbubble/
│   ├── bubbleGroup1_npz/
│   │   ├── bubbleGroup1_npz_1.tar.gz
│   │   ├── bubbleGroup1_npz_2.tar.gz
│   │   ├── .
│   │   ├── .
│   │   ├── .
│   │   └── bubbleGroup1_npz_10.tar.gz
│   ├── bubbleGroup2_npz/
│   │   ├── bubbleGroup2_npz_1.tar.gz
│   │   ├── bubbleGroup2_npz_2.tar.gz
│   │   ├── .
│   │   ├── .
│   │   ├── .
│   │   └── bubbleGroup2_npz_10.tar.gz
.
.
.
│   ├── bubbleGroup5_npz/
│   │   ├── bubbleGroup5_npz_1.tar.gz
│   │   ├── bubbleGroup5_npz_2.tar.gz
│   │   ├── .
│   │   ├── .
│   │   ├── .
│   │   └── bubbleGroup5_npz_10.tar.gz
│       └── mpf_paper_dataset/
│           ├── d1X_test_pad_flat.npz
│           ├── d1X_train_pad_flat.npz
│           ├── d1Y_test_pad_flat.npz
│           ├── d1Y_train_pad_flat.npz
│           ├── d2X_test_pad_flat.npz
│           ├── d2X_train_pad_flat.npz
│           ├── d2Y_test_pad_flat.npz
│           ├── d2Y_train_pad_flat.npz
│           ├── ...
├── 2Ddrop/
│   ├── dropGroup1_npz/
│   │   ├── dropGroup1_npz_1.tar.gz
│   │   ├── dropGroup1_npz_2.tar.gz
│   │   ├── .
│   │   ├── .
│   │   ├── .
│   │   └── dropGroup1_npz_10.tar.gz
│   ├── dropGroup2_npz/
│   │   ├── dropGroup2_npz_1.tar.gz
│   │   ├── dropGroup2_npz_2.tar.gz
│   │   ├── .
│   │   ├── .
│   │   ├── .
│   │   └── dropGroup2_npz_10.tar.gz
.
.
.
│   ├── dropGroup5_npz/
│   │   ├── dropGroup5_npz_1.tar.gz
│   │   ├── dropGroup5_npz_2.tar.gz
│   │   ├── .
│   │   ├── .
│   │   ├── .
│   │   └── dropGroup5_npz_10.tar.gz
├── 3Dbubble/
│   ├── split/
│   │   ├── 3DbubbleGroup1_npz_aa
│   │   ├── 3DbubbleGroup1_npz_ab
│   │   ├── .
│   │   ├── .
│   │   ├── .
│   │   └── 3DbubbleGroup1_npz_af
│   │   ├── 3DbubbleGroup2_npz_aa
│   │   ├── 3DbubbleGroup2_npz_ab
│   │   ├── .
│   │   ├── .
│   │   ├── .
│   │   └──  3DbubbleGroup2_npz_af
.
.
.
│   │   ├── 3DbubbleGroup5_npz_aa
│   │   ├── 3DbubbleGroup5_npz_ab
│   │   ├── .
│   │   ├── .
│   │   ├── .
│   │   └──  3DbubbleGroup5_npz_af
├── 3Ddrop/
│   ├── split/
│   │   ├── 3DdropGroup1_npz_aa
│   │   ├── 3DdropGroup1_npz_ab
│   │   ├── .
│   │   ├── .
│   │   ├── .
│   │   └── 3DdropGroup1_npz_af
│   │   ├── 3DdropGroup2_npz_aa
│   │   ├── 3DdropGroup2_npz_ab
│   │   ├── .
│   │   ├── .
│   │   ├── .
│   │   └──  3DdropGroup2_npz_af
.
.
.
│   │   ├── 3DdropGroup5_npz_aa
│   │   ├── 3DdropGroup5_npz_ab
│   │   ├── .
│   │   ├── .
│   │   ├── .
│   │   └──  3DdropGroup5_npz_af
├── Sample Dataset/
│   ├── 2D/
│   │   ├── bubble/
│   │   │   ├── Sample1/
│   │   │   │   |── X_bubble_padded_dataset_1.npz
│   │   │   │   |── Y_bubble_padded_dataset_1.npz
│   │   │   ├── Sample2/
│   │   │   │   |── X_bubble_padded_dataset_2.npz
│   │   │   │   |── Y_bubble_padded_dataset_2.npz
.
.
.
│   │   │   ├── Sample10/
│   │   │   │   |── X_bubble_padded_dataset_10.npz
│   │   │   │   |── Y_bubble_padded_dataset_10.npz
│   │   ├── drop/
│   │   │   ├── Sample1/
│   │   │   │   |── X_drop_padded_dataset_1.npz
│   │   │   │   |── Y_drop_padded_dataset_1.npz
│   │   │   ├── Sample2/
│   │   │   │   |── X_drop_padded_dataset_2.npz
│   │   │   │   |── Y_drop_padded_dataset_2.npz
.
.
.
│   │   │   ├── Sample10/
│   │   │   │   |── X_drop_padded_dataset_10.npz
│   │   │   │   |── Y_drop_padded_dataset_10.npz
│   ├── 3D/
│   │   ├── bubble/
│   │   │   ├── Sample1/
│   │   │   │   |── X_bubble_padded_dataset_3d_1.npz
│   │   │   │   |── Y_bubble_padded_dataset_3d_1.npz
│   │   │   ├── Sample2/
│   │   │   │   |── X_bubble_padded_dataset_3d_2.npz
│   │   │   │   |── Y_bubble_padded_dataset_3d_2.npz
.
.
.
│   │   │   ├── Sample10/
│   │   │   │   |── X_bubble_padded_dataset_3d_10.npz
│   │   │   │   |── Y_bubble_padded_dataset_3d_10.npz
│   │   ├── drop/
│   │   │   ├── Sample1/
│   │   │   │   |── X_drop_padded_dataset_3d_1.npz
│   │   │   │   |── Y_drop_padded_dataset_3d_1.npz
│   │   │   ├── Sample2/
│   │   │   │   |── X_drop_padded_dataset_3d_2.npz
│   │   │   │   |── Y_drop_padded_dataset_3d_2.npz
.
.
.
│   │   │   ├── Sample10/
│   │   │   │   |── X_drop_padded_dataset_3d_10.npz
│   │   │   │   |── Y_drop_padded_dataset_3d_10.npz
├── README.md
├── .gitattributes
└── info.txt
```

**Citation**
If you find this dataset useful in your research, please consider citing our paper as follows: 
```
@article{shadkhah2024MPFBench,
        title = "FlowBench: A Large Scale Benchmark for Flow Simulation             over Complex Geometries",
        author = "Shadkhah, Mehdi and Tali, Ronak and Rabeh, Ali and Yang, Cheng-Hau and Upadhyaya, Abhisek and Krishnamurthy, Adarsh and Hegde, Chinmay and Balu, Aditya and Ganapathysubramanian, Baskar"
        year = "2024"
}
```