Module monk.pytorch.datasets.paths

Functions

def set_dataset_test_path(system_dict, path, path_to_csv, delimiter)

Set dataset train path

Args

system_dict : dict

System dictionary containing all the variables

path : str, list

Dataset folder path 1) String : For dataset with no validation set 2) List: For dataset with validation set in order [train_set, val_set]

path_to_csv : str, list

Path to csv pointing to images

delimiter : str

Delimiter for the csv path provided

Returns

dict

Updated System dictionary

Expand source code

def set_dataset_test_path(system_dict,  path, path_to_csv, delimiter):
    '''
    Set dataset train path

    Args:
        system_dict (dict): System dictionary containing all the variables
        path (str, list): Dataset folder path
                          1) String : For dataset with no validation set
                          2) List: For dataset with validation set in order [train_set, val_set]
        path_to_csv (str, list): Path to csv pointing to images
        delimiter (str): Delimiter for the csv path provided

    Returns:
        dict: Updated System dictionary
    '''
    dataset_test_type = None;
    dataset_test_path = None;
    csv_test = None;
    if(path_to_csv):
        csv_test = path_to_csv;
        dataset_test_path = path;
        dataset_test_type = "csv";
    else:
        dataset_test_path = path;
        dataset_test_type = "foldered";
    
    system_dict["dataset"]["test_path"] = dataset_test_path;
    system_dict["dataset"]["csv_test"] = csv_test;
    system_dict["dataset"]["params"]["test_delimiter"] = delimiter;
    system_dict["dataset"]["params"]["dataset_test_type"] = dataset_test_type;


    return system_dict;

def set_dataset_train_path(system_dict, path, split, path_to_csv, delimiter)

Set dataset train path

Args

system_dict : dict

System dictionary containing all the variables

path : str, list

Dataset folder path 1) String : For dataset with no validation set 2) List: For dataset with validation set in order [train_set, val_set]

split : float

Indicating train validation split Division happens as follows: train - total dataset * split * 100 val - total dataset * (1-split) * 100

path_to_csv : str, list

Path to csv pointing to images

delimiter : str

Delimiter for the csv path provided

Returns

dict

Updated System dictionary

Expand source code

def set_dataset_train_path(system_dict, path, split, path_to_csv, delimiter):
    '''
    Set dataset train path

    Args:
        system_dict (dict): System dictionary containing all the variables
        path (str, list): Dataset folder path
                          1) String : For dataset with no validation set
                          2) List: For dataset with validation set in order [train_set, val_set]
        split (float): Indicating train validation split
                       Division happens as follows:
                           train - total dataset * split * 100
                           val   - total dataset * (1-split) * 100 
        path_to_csv (str, list): Path to csv pointing to images
        delimiter (str): Delimiter for the csv path provided

    Returns:
        dict: Updated System dictionary
    '''
    dataset_type = None;
    dataset_train_path = None;
    dataset_val_path = None;
    csv_train = None;
    csv_val = None;
    train_val_split = None;
    if(path_to_csv):
        if(type(path) == str):
            dataset_type = "csv_train";
            csv_train = path_to_csv;
            dataset_train_path = path;
            train_val_split = split;
        elif(type(path) == list):
            dataset_type = "csv_train-val";
            csv_train = path_to_csv[0];
            csv_val = path_to_csv[1];
            dataset_train_path = path[0];
            dataset_val_path = path[1];
            train_val_split = None;
    else:
        if(type(path) == str):
            dataset_type = "train";
            dataset_train_path = path;
            train_val_split = split;
        elif(type(path) == list):
            dataset_type = "train-val";
            dataset_train_path = path[0];
            dataset_val_path = path[1];
            train_val_split = None;

    system_dict["dataset"]["dataset_type"] = dataset_type;
    system_dict["dataset"]["train_path"] = dataset_train_path;
    system_dict["dataset"]["val_path"] = dataset_val_path;
    system_dict["dataset"]["csv_train"] = csv_train;
    system_dict["dataset"]["csv_val"] = csv_val;
    system_dict["dataset"]["params"]["train_val_split"] = train_val_split;
    system_dict["dataset"]["params"]["delimiter"] = delimiter;

    return system_dict;