Mercurial > ift6266
changeset 175:224321bf043a
Define the ocr dataset and use the existing split for nist.
| author | Arnaud Bergeron <abergeron@gmail.com> |
|---|---|
| date | Sat, 27 Feb 2010 13:56:14 -0500 |
| parents | ff26436d42d6 |
| children | d6672a7daea5 |
| files | datasets/__init__.py datasets/defs.py datasets/nist.py |
| diffstat | 3 files changed, 37 insertions(+), 24 deletions(-) [+] |
line wrap: on
line diff
--- a/datasets/__init__.py Sat Feb 27 12:18:26 2010 -0500 +++ b/datasets/__init__.py Sat Feb 27 13:56:14 2010 -0500 @@ -1,1 +1,2 @@ -from nist import * +from defs import * +
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/datasets/defs.py Sat Feb 27 13:56:14 2010 -0500 @@ -0,0 +1,35 @@ +__all__ = ['nist_digits', 'nist_lower', 'nist_upper', 'nist_all'] + +from ftfile import FTDataSet + +NIST_PATH = '/data/lisa/data/nist/by_class/' +DATA_PATH = '/data/lisa/data/ift6266h10/' + +nist_digits = FTDataSet(train_data = [NIST_PATH+'digits/digits_train_data.ft'], + train_lbl = [NIST_PATH+'digits/digits_train_labels.ft'], + test_data = [NIST_PATH+'digits/digits_test_data.ft'], + test_lbl = [NIST_PATH+'digits/digits_test_labels.ft']) +nist_lower = FTDataSet(train_data = [NIST_PATH+'lower/lower_train_data.ft'], + train_lbl = [NIST_PATH+'lower/lower_train_labels.ft'], + test_data = [NIST_PATH+'lower/lower_test_data.ft'], + test_lbl = [NIST_PATH+'lower/lower_test_labels.ft']) +nist_upper = FTDataSet(train_data = [NIST_PATH+'upper/upper_train_data.ft'], + train_lbl = [NIST_PATH+'upper/upper_train_labels.ft'], + test_data = [NIST_PATH+'upper/upper_test_data.ft'], + test_lbl = [NIST_PATH+'upper/upper_test_labels.ft']) + + + +nist_all = FTDataSet(train_data = [DATA_PATH+'train_data.ft'], + train_lbl = [DATA_PATH+'train_labels.ft'], + test_data = [DATA_PATH+'test_data.ft'], + test_lbl = [DATA_PATH+'test_labels.ft'], + valid_data = [DATA_PATH+'valid_data.ft'], + valid_lbl = [DATA_PATH+'valid_labels.ft']) + +ocr = FTDataSet(train_data = [DATA_PATH+'ocr_train_data.ft'], + train_lbl = [DATA_PATH+'ocr_train_labels.ft'], + test_data = [DATA_PATH+'ocr_test_data.ft'], + test_lbl = [DATA_PATH+'ocr_test_labels.ft'], + valid_data = [DATA_PATH+'ocr_valid_data.ft'], + valid_lbl = [DATA_PATH+'ocr_valid_labels.ft'])
--- a/datasets/nist.py Sat Feb 27 12:18:26 2010 -0500 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,23 +0,0 @@ -__all__ = ['nist_digits', 'nist_lower', 'nist_upper', 'nist_all'] - -from ftfile import FTDataSet - -PATH = '/data/lisa/data/nist/by_class/' - -nist_digits = FTDataSet(train_data = [PATH+'digits/digits_train_data.ft'], - train_lbl = [PATH+'digits/digits_train_labels.ft'], - test_data = [PATH+'digits/digits_test_data.ft'], - test_lbl = [PATH+'digits/digits_test_labels.ft']) -nist_lower = FTDataSet(train_data = [PATH+'lower/lower_train_data.ft'], - train_lbl = [PATH+'lower/lower_train_labels.ft'], - test_data = [PATH+'lower/lower_test_data.ft'], - test_lbl = [PATH+'lower/lower_test_labels.ft']) -nist_upper = FTDataSet(train_data = [PATH+'upper/upper_train_data.ft'], - train_lbl = [PATH+'upper/upper_train_labels.ft'], - test_data = [PATH+'upper/upper_test_data.ft'], - test_lbl = [PATH+'upper/upper_test_labels.ft']) -nist_all = FTDataSet(train_data = [PATH+'all/all_train_data.ft'], - train_lbl = [PATH+'all/all_train_labels.ft'], - test_data = [PATH+'all/all_test_data.ft'], - test_lbl = [PATH+'all/all_test_labels.ft']) -