def train_model(classifier, train_csv, test_csv, model_file):
train_set = digit_io.read_from_csv(train_csv)
test_set = digit_io.read_from_csv(test_csv)
assert train_set.ndim == 2
assert test_set.ndim == 2
data_train, target_train = train_set[0::, 1::], train_set[0::, 0]
test_train, target_test = test_set[0::, 1::], test_set[0::, 0]
# classifier = RandomForestClassifier(n_estimators=100)
learn_model(classifier, data_train, test_train, target_train, target_test, model_file)
def train_model(classifier, train_csv, test_csv, model_file):
train_set = digit_io.read_from_csv(train_csv)
test_set = digit_io.read_from_csv(test_csv)
assert train_set.ndim == 2
assert test_set.ndim == 2
data_train, target_train = train_set[0::, 1::], train_set[0::, 0]
test_train, target_test = test_set[0::, 1::], test_set[0::, 0]
# classifier = RandomForestClassifier(n_estimators=100)
learn_model(classifier, data_train, test_train, target_train, target_test,
model_file)
def split_entity_dataset():
train_csv_path = "../data/train.csv"
data_set = digit_io.read_from_csv(train_csv_path)
train_set, test_set = data_split(data_set)
digit_io.write_to_csv(train_set, '../data/training/train.csv')
digit_io.write_to_csv(test_set, '../data/training/test.csv')
def split_entity_dataset():
train_csv_path = "../data/train.csv"
data_set = digit_io.read_from_csv(train_csv_path)
train_set, test_set = data_split(data_set)
digit_io.write_to_csv(train_set, '../data/training/train.csv')
digit_io.write_to_csv(test_set, '../data/training/test.csv')
def predict(model_file, test_file, filename):
classifier = load_model(model_file)
test = digit_io.read_from_csv(test_file)
predicted = classifier.predict(test)
if filename != None:
# digit_io.write_to_csv(predicted, filename, header=None)
digit_io.write_predicted_result(predicted, filename)
else:
print predicted
def extract_small_dataset():
train_csv_path = "../data/train.csv"
data_set = digit_io.read_from_csv(train_csv_path)
train_set, test_set = data_split(data_set)
small_train, small_test = data_split(test_set)
digit_io.write_to_csv(small_train, "../data/training/small_train.csv")
digit_io.write_to_csv(small_test, "../data/training/small_test.csv")
def predict(model_file, test_file, filename) :
classifier = load_model(model_file)
test = digit_io.read_from_csv(test_file)
predicted = classifier.predict(test)
if filename != None:
# digit_io.write_to_csv(predicted, filename, header=None)
digit_io.write_predicted_result(predicted, filename)
else:
print predicted
def extract_small_dataset():
train_csv_path = "../data/train.csv"
data_set = digit_io.read_from_csv(train_csv_path)
train_set, test_set = data_split(data_set)
small_train, small_test = data_split(test_set)
digit_io.write_to_csv(small_train, "../data/training/small_train.csv")
digit_io.write_to_csv(small_test, "../data/training/small_test.csv")
