notwa/mnists
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

remove incomplete npz functionality

Browse Source
This commit is contained in:
Connor Olding 2018-03-24 11:16:26 +01:00
parent c7da7e5fbf
commit 6efccb6b20
2 changed files with 20 additions and 30 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
TODO
View File

@ -2,8 +2,6 @@ TODO
* finish writing README
* finish npz functionality
* document prepare() function
* support python 3.5
@ -17,3 +15,5 @@ TODO
* document everything
* test everything
* consider npz functionality if it reduces preprocessing time

46
mnists/__init__.py
View File

@ -5,8 +5,6 @@
__version__ = "0.2.1"
# NOTE: npz functionality is incomplete.
import array
import gzip
import hashlib
@ -29,29 +27,29 @@ output_directory = os.path.join(home, ".mnist")
webhost = "https://eaguru.guru/mnist/"
def _make_meta(npz, train_part="train", test_part="t10k", prefix=""):
def _make_meta(train_part="train", test_part="t10k", prefix=""):
images_suffix = "-images-idx3-ubyte.gz"
labels_suffix = "-labels-idx1-ubyte.gz"
return (prefix, npz,
return (prefix,
train_part + images_suffix,
train_part + labels_suffix,
test_part + images_suffix,
test_part + labels_suffix)
def _emnist_meta(npz, name):
return _make_meta(npz, name + "-train", name + "-test", prefix="emnist")
def _emnist_meta(name):
return _make_meta(name + "-train", name + "-test", prefix="emnist")
metadata = dict(
emnist_balanced=_emnist_meta("emnist_balanced.npz", "emnist-balanced"),
emnist_byclass=_emnist_meta("emnist_byclass.npz", "emnist-byclass"),
emnist_bymerge=_emnist_meta("emnist_bymerge.npz", "emnist-bymerge"),
emnist_digits=_emnist_meta("emnist_digits.npz", "emnist-digits"),
emnist_letters=_emnist_meta("emnist_letters.npz", "emnist-letters"),
emnist_mnist=_emnist_meta("emnist_mnist.npz", "emnist-mnist"),
fashion_mnist=_make_meta("fashion_mnist.npz", prefix="fashion-mnist"),
mnist=_make_meta("mnist.npz", prefix="mnist"),
emnist_balanced=_emnist_meta("emnist-balanced"),
emnist_byclass=_emnist_meta("emnist-byclass"),
emnist_bymerge=_emnist_meta("emnist-bymerge"),
emnist_digits=_emnist_meta("emnist-digits"),
emnist_letters=_emnist_meta("emnist-letters"),
emnist_mnist=_emnist_meta("emnist-mnist"),
fashion_mnist=_make_meta(prefix="fashion-mnist"),
mnist=_make_meta(prefix="mnist"),
)
@ -160,7 +158,7 @@ def prepare(dataset="mnist", return_floats=True, return_onehot=True,
meta = metadata[dataset]
prefix, names = meta[0], meta[1:]
names = [os.path.join(prefix, name) for name in names]
npz, train_images, train_labels, test_images, test_labels = names
train_images, train_labels, test_images, test_labels = names
images_and_labels = names[1:]
logger.debug("Filenames chosen for %s: %s, %s, %s, %s",
@ -169,25 +167,17 @@ def prepare(dataset="mnist", return_floats=True, return_onehot=True,
make_directories()
existing = [os.path.isfile(construct_path(name)) for name in names]
npz_existing, gz_existing = existing[0], existing[1:]
gz_existing = existing[0], existing[1:]
for name in images_and_labels:
download(name)
if check_integrity:
validate(name)
if npz_existing:
logger.info("Loading npz file %s", npz)
with np.load(construct_path(npz)):
train_images_data, train_labels_data, \
test_images_data, test_labels_data = \
f["train_images"], f["train_labels"], \
f["test_images"], f["test_labels"]
else:
train_images_data, train_labels_data = load(
construct_path(train_images), construct_path(train_labels))
test_images_data, test_labels_data = load(
construct_path(test_images), construct_path(test_labels))
train_images_data, train_labels_data = load(
construct_path(train_images), construct_path(train_labels))
test_images_data, test_labels_data = load(
construct_path(test_images), construct_path(test_labels))
# correct the orientation of emnist images.
if prefix == "emnist":