ex_time_mc.py

from api import API
from disaggregate import ADAE, DAE, Seq2Point, Seq2Seq, WindowGRU, RNN
import warnings

warnings.filterwarnings("ignore")

path = 'D:/workspace/nilm/data/redd_data.h5'
# path = 'D:/workspace/nilm/code/databank/redd_data.h5'

DEBUG = False
TEST = False


def generate_method(debug, test):
    if debug:
        method = {
            'DAE': DAE({'save-model-path': 'DAE', 'pretrained-model-path': None, 'n_epochs': 1, 'batch_size': 256}),
            # 'RNN': RNN({'save-model-path': 'RNN', 'pretrained-model-path': None, 'n_epochs': 1, 'batch_size': 256}),
            # 'Seq2Point': Seq2Point({'save-model-path': 'Seq2Point', 'pretrained-model-path': None, 'n_epochs': 1, 'batch_size': 256}),
            # 'Seq2Seq': Seq2Seq({'save-model-path': 'Seq2Seq', 'pretrained-model-path': None, 'n_epochs': 1, 'batch_size': 256}),
            # 'GRU': WindowGRU({'save-model-path': 'GRU', 'pretrained-model-path': None, 'n_epochs': 1, 'batch_size': 256}),
        }
    else:
        method = {
            'DAE': DAE({'save-model-path': 'DAE', 'pretrained-model-path': None}),
            'RNN': RNN({'save-model-path': 'RNN', 'pretrained-model-path': None}),
            'Seq2Point': Seq2Point({'save-model-path': 'Seq2Point', 'pretrained-model-path': None}),
            'Seq2Seq': Seq2Seq({'save-model-path': 'Seq2Seq', 'pretrained-model-path': None}),
            'GRU': WindowGRU({'save-model-path': 'GRU', 'pretrained-model-path': None}),
        }
    if test:
        method = {
            'DAE': DAE({'save-model-path': 'DAE', 'pretrained-model-path': 'DAE', 'batch_size': 256}),
            # 'RNN': RNN({'save-model-path': 'RNN', 'pretrained-model-path': 'RNN', 'batch_size': 256}),
            # 'Seq2Point': Seq2Point(
            #     {'save-model-path': 'Seq2Point', 'pretrained-model-path': 'Seq2Point', 'batch_size': 256}),
            # 'Seq2Seq': Seq2Seq({'save-model-path': 'Seq2Seq', 'pretrained-model-path': 'Seq2Seq', 'batch_size': 256}),
            # 'GRU': WindowGRU({'save-model-path': 'GRU', 'pretrained-model-path': 'GRU', 'batch_size': 256}),
        }
    return method

time_config = {
    'train': {
        1: {
            'start_time': '2011-04-18',
            'end_time': '2011-05-07'
        },
        2: {
            'start_time': '2011-04-17',
            'end_time': '2011-04-25'
        },
        3: {
            'start_time': '2011-04-16',
            'end_time': '2011-04-27'
        },

        4: {
            'start_time': '2011-04-16',
            'end_time': '2011-05-22'
        },
        6: {
            'start_time': '2011-04-16',
            'end_time': '2011-06-09'
        }
    },
    'test': {
        1: {
            'start_time': '2011-05-07',
            'end_time': '2011-05-24'
        },
        2: {
            'start_time': '2011-04-25',
            'end_time': '2011-05-22'
        },
        3: {
            'start_time': '2011-04-27',
            'end_time': '2011-05-30'
        },

        4: {
            'start_time': '2011-05-22',
            'end_time': '2011-06-03'
        },
        6: {
            'start_time': '2011-06-09',
            'end_time': '2011-06-13'
        }
    }
}

method = generate_method(DEBUG, TEST)

ex_train_microwave = {

    'power': {
        'mains': ['apparent', 'active'],
        'appliance': ['apparent', 'active']
    },
    'sample_rate': 6,

    'appliances': ['microwave'],
    'methods': method,
    'isState': False,
    'train': {
        'datasets': {

            'redd': {
                'path': path,
                'buildings': {
                    1: time_config['train'][1],
                    2: time_config['train'][2],
                    3: time_config['train'][3],
                }

            }
        }
    },

    'test': {
        'datasets': {
            'redd': {
                'path': path,
                'buildings': {
                    1: time_config['test'][1],
                    # 2: time_config['test'][2],
                    # 3: time_config['test'][3],
                }
            }
        },
    },
}
API(ex_train_microwave)