# RFI_Net

**Repository Path**: AstroTJU/RFI_Net

## Basic Information

- **Project Name**: RFI_Net
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: GPL-3.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2020-08-02
- **Last Updated**: 2020-12-19

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# RFI-Net

## Description
This repositories is for the Radio frequency interference (RFI) detection

## Software Architecture
In the folder data-set, the file data_preprocessing.py is to packed the data into tfrecord for train and validation. The test data is packed as hdf5 file.

In the folder RFI-Net, the rfi-net.py is where the model is constructed and train, validate and test.

## Instructions

1.  import rfi_net
```python
import sys
sys.path.append("..")
from RFI_Net.RFI_Net.rfi_net import RFI_Net

net = RFI_Net()
```

2.  set arguments and train a model
```python
net.train(train_file_path="train_set.tfrecords", model_file_path="data_set/saved_models", log_path="data_set/log")
```

3.  detect
```python
net.detect(detect_batch_size=16, detect_set_size=4096, need_detect_file_path="data_set/to_be_detect.h5", detect_result_path="data_set/detect_result.h5", model_file_path="data_set/saved_models")
```

## Detection result
The detection result is show as below:

![image](data_set/pictures/detection_result.jpg)

The paper of RFI-Net has been published in Monthly Notices of the Royal Astronomical Society (please see [Deep residual detection of radio frequency interference for FAST](https://doi.org/10.1093/mnras/stz3521 "RFI-Net for RFI detection"))

## Contribution are Wellcome

1.  Fork the repository
2.  Create dev/xxx branch
3.  Commit your code
4.  Create Pull Request