prediction.py

#!/usr/bin/env python3

"""
Inference using provided pre-trained model


Editor: Marshall Xu
Last edited: 18/10/2023
"""

import os
import config.pred_config as pred_config
from utils import preprocess_procedure, make_prediction

args = pred_config.pred_parser.parse_args()

ds_path = args.ds_path # path to original data
ps_path = args.ps_path # path to preprocessed data
out_path = args.out_path # path to infered data
if os.path.exists(out_path) == False:
    print(f"{out_path} does not exist.")
    os.mkdir(out_path)
    print(f"{out_path} has been created!")

prep_mode = args.prep_mode # preprocessing mode
# when the preprocess is skipped, 
# directly take the raw data for inference
if prep_mode == 4:
    ps_path = ds_path

model_type = args.mo # model type
in_chan = args.ic # input channel
ou_chan = args.oc # output channel
fil_num = args.fil # number of filters

threshold_vector = [args.thresh, args.cc]
pretrained_model = args.pretrained # path to pretrained model

if __name__ == "__main__":

    print("Prediction session will start shortly..")

    # preprocess procedure
    preprocess_procedure(ds_path, ps_path, prep_mode)

    # make prediction
    make_prediction(model_type, in_chan, ou_chan,
                    fil_num, ps_path, out_path,
                    threshold_vector[0], threshold_vector[1], pretrained_model,
                    mip_flag=True)