def sample(self, n_per_class=10, preload=False):
import random
tran = listvalues(self.trans)[0]
splt = listvalues(self.splits)[0]
the_sample = []
for cls in listvalues(self.classes):
cls: ImageDatasetClass
folder = cls.folder(splt, tran)
im_paths = folder.paths
num = len(im_paths)
assert num > n_per_class
already_took = []
n = 0
while n < n_per_class:
i = random.randrange(0, num)
if i in already_took:
continue
else:
the_im = ImageDatasetImage(File(im_paths[i]), self, tran,
splt, cls)
if preload:
the_im.load()
the_sample.append(the_im)
n += 1
already_took += [i]
return the_sample
def violin(self):
fd = PlotData(
y=[y.tolist() for y in listvalues(self.data)],
item_type='violin',
**self._common(),
)
return fd
def bar(self):
means = {k: np.nanmean(v) for k, v in self.data.items()}
fd = PlotData(
y=listvalues(means),
item_type='bar',
**self._common(),
)
return fd
def __init__(self, name):
self.folder = OM_IM_DATA_ROOT[name]
self.metadata_file = self.folder['metadata.json']
self.trans = {_RAW: ImageDatasetTransformation(self, _RAW)}
self.splits = {
n: ImageDatasetSplit(self, n)
for n in [_TRAIN, _TEST, _EVAL]
}
if not ismac():
self.classes = ClassSet([
ImageDatasetClass(name=File(class_folder).name,
index=i,
dataset=self) for i, class_folder in esorted(
listvalues(self.splits)[0].folder(
listvalues(self.trans)[0]).paths)
])
if not ismac():
self.verify()
def vals(self) -> li:
return li(listvalues(self))
def dict_to_table(d):
from mlib.math import sigfig_if_num
return [[''] + [k for k in listvalues(d)[0].keys()]] + [
[k] + listvalues(v).map(sigfig_if_num) for k, v in d.items()
]