Created
June 15, 2018 10:13
-
-
Save danlester/f7cdbc741cf06317ff3870c52bb3474a to your computer and use it in GitHub Desktop.
Tensorflow Datasets API for processing CSV string column into int tensor
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
CSV_COLUMNS = ['p{}'.format(i) for i in range(30)] + ['Image'] | |
CSV_COLUMN_DEFAULTS = [[0.0]]*30 + [['']] | |
def parse_csv(rows_string_tensor): | |
columns = tf.decode_csv(rows_string_tensor, record_defaults=CSV_COLUMN_DEFAULTS) | |
raw_features = dict(zip(CSV_COLUMNS, columns)) | |
image_str_array_sparse = tf.string_split([raw_features['Image']]) | |
image_str_array = tf.sparse_to_dense(image_str_array_sparse.indices, image_str_array_sparse.dense_shape, image_str_array_sparse.values, '') | |
image_str_array = image_str_array[0] | |
raw_features['Image'] = tf.string_to_number(image_str_array) | |
return raw_features | |
def datasetload(csvfile, test=False): | |
return tf.data.TextLineDataset([csvfile]).skip(1).map(parse_csv) | |
features_and_labels = datasetload('training.csv') |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment