def test_ExampleConcatenatedDataset_slicing():
D1 = DebugDataset(size=10)
D2 = DebugDataset(size=10, offset=1)
D3 = DebugDataset(size=10, offset=2)
D4 = DebugDataset(size=10, offset=3)
E = ExampleConcatenatedDataset(D1, D2, D3, D4)
E.set_example_pars(start=1, step=2)
assert len(E) == 10
d = E[2]
assert d == {"val": [2 + 1, 2 + 3], "index_": 2, "other": [2 + 1, 2 + 3]}
assert len(E.labels["label1"]) == 10
assert np.all(E.labels["label1"] == [[i + 1, i + 3] for i in range(10)])
E.set_example_pars(start=0, stop=-1, step=2)
assert len(E) == 10
d = E[2]
assert d == {"val": [2, 2 + 2], "index_": 2, "other": [2, 2 + 2]}
assert len(E.labels["label1"]) == 10
assert np.all(E.labels["label1"] == [[i, i + 2] for i in range(10)])
E.set_example_pars(start=1, stop=-1, step=2)
assert len(E) == 10
d = E[2]
assert d == {"val": [2 + 1], "index_": 2, "other": [2 + 1]}
assert len(E.labels["label1"]) == 10
assert np.all(E.labels["label1"] == [[i + 1] for i in range(10)])
def test_unseq_vanilla():
D1 = DebugDataset(size=10)
D2 = DebugDataset(size=10)
D = D1 + D2
S = SequenceDataset(D, 3, fid_key="label1")
U = UnSequenceDataset(S)
print(np.shape(U.labels["label1"]))
assert len(U) == len(S) * 3
assert len(U.labels["label1"]) == len(S.labels["label1"]) * 3
du = U[[0, 1, 2]]
ds = S[0]
for k in ds.keys():
if k == "index_":
continue
for i in range(3):
print(k, i)
val_u = du[i][k]
val_s = ds[k][i]
assert val_s == val_u
def test_ExampleConcatenatedDataset_step():
D1 = DebugDataset(size=10)
D2 = DebugDataset(size=10)
E = ExampleConcatenatedDataset(D1, D2)
E.set_example_pars(step=2)
assert len(E) == 10
d = E[2]
assert d == {"val": [2], "index_": 2, "other": [2]}
assert len(E.labels["label1"]) == 10
assert np.all(E.labels["label1"] == [[i, i] for i in range(10)])
D3 = DebugDataset(size=20)
with pytest.raises(AssertionError):
ExampleConcatenatedDataset(D1, D3)
D4 = DebugDataset(size=10)
D5 = DebugDataset(size=10)
E = ExampleConcatenatedDataset(D1, D2, D4, D5)
E.set_example_pars(step=2)
assert len(E) == 10
d = E[2]
assert d == {"val": [2, 2], "index_": 2, "other": [2, 2]}
assert len(E.labels["label1"]) == 10
assert np.all(E.labels["label1"] == [[i, i] for i in range(10)])
def test_sequence_dset_vanilla():
D1 = DebugDataset(size=10)
D2 = DebugDataset(size=10)
D3 = DebugDataset(size=10)
D = D1 + D2 + D3
S = SequenceDataset(D, 3, fid_key="label1")
assert len(S) == (3 * (10 - 2))
assert len(S.labels["label1"]) == (3 * (10 - 2))
d = S[0]
ref = {
"val": [0, 1, 2],
"other": [0, 1, 2],
"index_": 0,
"dataset_index_": [0, 0, 0],
}
assert d == ref
l = np.array(S.labels["label1"][:3])
refl = np.array([[0, 1, 2], [1, 2, 3], [2, 3, 4]])
assert np.all(l == refl)
def test_dset_mxin_ops():
"""Basically test the ConcatenatedDataset"""
class MyDset(DatasetMixin):
def __init__(self):
self.data = DebugDataset(size=10)
D1 = DebugDataset(size=10)
D2 = DebugDataset(size=10)
D3 = D1 + D2
assert len(D3) == 20
D3[13]
D3 = D2 + D1
assert len(D3) == 20
D3[13]
D4 = 3 * D1
assert len(D4) == 30
D4[13]
D4[23]
D4 = D1 * 3
assert len(D4) == 30
D4[13]
D4[23]
def test_seq_wrong_fid_dtype():
D1 = DebugDataset(size=10)
D2 = DebugDataset(size=10)
D = D1 + D2
D.labels["label1"] = np.array(D.labels["label1"], dtype=float)
with pytest.raises(TypeError):
S = SequenceDataset(D, 3, fid_key="label1")
def test_seq_offset_fid_strat_reset():
D1 = DebugDataset(size=10)
D2 = DebugDataset(size=12, offset=3)
D = D1 + D2
S = SequenceDataset(D, 3, fid_key="label1", strategy="reset")
assert len(S) == (len(D1) - 2 + len(D2) - 2)
def test_seq_wrong_fid_len():
D1 = DebugDataset(size=10)
D2 = DebugDataset(size=10)
D = D1 + D2
D.labels["label1"] = np.array(D.labels["label1"])[:10]
with pytest.raises(ValueError):
S = SequenceDataset(D, 3, fid_key="label1")
def test_DisjunctExampleConcatenatedDataset_not_same_length():
D1 = DebugDataset(size=10)
D2 = DebugDataset(size=20)
with pytest.raises(AssertionError):
D = DisjunctExampleConcatenatedDataset(D1, D2, disjunct=False, same_length=True)
D = DisjunctExampleConcatenatedDataset(D1, D2, disjunct=False, same_length=False)
assert len(D) == 10
def test_DisjunctExampleConcatenatedDataset_not_disjunct():
D1 = DebugDataset(size=10, offset=0)
D2 = DebugDataset(size=10, offset=10)
D = DisjunctExampleConcatenatedDataset(D1, D2, disjunct=False)
assert len(D) == 10
assert list(D.labels.keys()) == ["label1", "label2"]
assert D.labels == D2.labels
assert D[2] == {"val": 12, "index_": 2, "other": 12}
assert D[2] == D2[2]
def test_sequence_dset_offset_fid():
D1 = DebugDataset(size=10)
D2 = DebugDataset(size=10)
D3 = DebugDataset(size=10, offset=3)
D4 = DebugDataset(size=10, offset=3)
D = D1 + D2 + D3 + D4
labels = np.array(D.labels["label1"])
diffs = labels[1:] - labels[:-1]
idxs = np.where(diffs <= 0)[0] + 1
with pytest.raises(ValueError):
S = SequenceDataset(D, 3, fid_key="label1")
def test_get_seq():
config = {
"dataset": "edflow.data.believers.sequence.getSeqDataset",
"seqdataset": {
"dataset": "edflow.debug.ConfigDebugDataset",
"length": 3,
"step": 1,
"fid_key": "label1",
},
"size": 10,
}
S1 = getSeqDataset(config)
D = DebugDataset(size=10)
S2 = SequenceDataset(D, 3, fid_key="label1")
assert len(S1) == len(S2)
assert len(S1.labels["label1"]) == len(S2.labels["label1"])
assert len(S1) == len(S1.labels["label1"])
s1 = S1[0]
s2 = S2[0]
assert s1 == s2
assert np.all(S1.labels["label1"] == S2.labels["label1"])
def test_label_dset():
D = DebugDataset(size=10)
L = LabelDataset(D)
assert L[0] == {"label1": 0, "label2": 0, "index_": 0, "base_index_": 0}
assert len(L) == 10
assert len(L.labels["label1"]) == 10
def test_processed_dset_no_update():
D = DebugDataset(size=10)
P = ProcessedDataset(D, lambda val, index_, other: {"val": val**2}, False)
assert len(D) == len(P)
assert len(D.labels["label1"]) == len(P.labels["label1"])
dp = P[0]
assert dp == {"val": 0, "index_": 0}
def test_processed_dset():
D = DebugDataset(size=10)
P = ProcessedDataset(D, lambda val, index_, other: {"val": val**2})
assert len(D) == len(P)
assert len(D.labels["label1"]) == len(P.labels["label1"])
dp = P[0]
ref = D[0]
ref["val"] = ref["val"]**2
assert dp == ref
def test_DisjunctExampleConcatenatedDataset():
D1 = DebugDataset(size=10, other_labels=False, other_ex_keys=False)
D2 = DebugDataset(size=10, other_labels=True, other_ex_keys=True)
D3 = DebugDataset(size=10, other_labels=True, other_ex_keys=False)
D = DisjunctExampleConcatenatedDataset(D1, D2)
assert len(D) == 10
d = D[2]
assert d == {"val": 2, "index_": 2, "other": 2, "val_other": 2, "other_other": 2}
assert list(D.labels.keys()) == ["label1", "label2", "label1_other", "label2_other"]
assert len(D.labels["label1"]) == 10
assert len(D.labels["label1_other"]) == 10
with pytest.raises(AssertionError):
DisjunctExampleConcatenatedDataset(D1, D1)
D = DisjunctExampleConcatenatedDataset(D1, D3)
with pytest.raises(AssertionError):
D[2]
def test_extra_labels():
D = DebugDataset(size=10)
D.append_labels = True
E = ExtraLabelsDataset(D, lambda dset, idx: {"new": idx})
de = E[0]
ref = D[0]
set_value(ref, "labels_/new", 0)
assert de == ref
assert "new" in E.labels
assert np.all(E.labels["new"] == np.arange(10))
def test_debug_dset():
D = DebugDataset(size=10)
print(
bcolors.WARNING + bcolors.BOLD +
"Dear hacker,\nShould you see this message you probably have changed "
"something in the DebugDataset code. Please consider that all "
"dataset test depend on this piece of code, so you want to be very "
"careful what you change.\n Yours sincerely, jhaux" + bcolors.ENDC)
assert len(D) == 10
assert len(D.labels["label1"]) == 10
assert D[0] == {"val": 0, "other": 0, "index_": 0}
assert np.all(D.labels["label1"] == np.arange(10))
def test_ConcatenatedDataset_balanced():
D1 = DebugDataset(size=10)
D2 = DebugDataset(size=20)
C = ConcatenatedDataset(D1, D2, balanced=True)
assert len(C) == 40
ref = D1[2]
ref["dataset_index_"] = 0
ref["index_"] = 2
d = C[2]
assert d == ref
ref["index_"] = 12
d = C[12]
assert d == ref
ref = D2[2]
ref["dataset_index_"] = 1
ref["index_"] = 22
d = C[22]
assert d == ref
ref = D2[12]
ref["dataset_index_"] = 1
ref["index_"] = 32
d = C[32]
assert d == ref
assert len(C.labels["label1"]) == 40
lref = D1.labels["label1"][2]
l = C.labels["label1"][2]
assert l == lref
l = C.labels["label1"][12]
assert l == lref
l = C.labels["label1"][12]
assert l == lref
def test_sub():
D = DebugDataset(10)
I = np.array([9, 1, 2, 4, 3, 5, 7, 6, 8, 0])
S = SubDataset(D, I)
ref0 = {"val": 9, "other": 9, "index_": 0}
ref2 = {"val": 2, "other": 2, "index_": 2}
ref6 = {"val": 7, "other": 7, "index_": 6}
assert S[0] == ref0
assert S[2] == ref2
assert S[6] == ref6
assert all(S.labels["label1"] == I)
def test_ConcatenatedDataset():
D = DebugDataset(size=10)
C = ConcatenatedDataset(D, D)
assert len(C) == 20
ref = D[2]
ref["dataset_index_"] = 0
d = C[2]
assert d == ref
d = C[12]
ref["dataset_index_"] = 1
ref["index_"] = 12
assert d == ref
assert len(C.labels["label1"]) == 20
lref = D.labels["label1"][2]
l = C.labels["label1"][2]
assert l == lref
l = C.labels["label1"][12]
assert l == lref
if 'model_output.csv' in root:
root = root[:-len('model_output.csv')]
savename_score = os.path.join(save_dir, 'score.txt')
#savename_std = os.path.join(save_dir, 'std.txt')
fid_score = np.array(fids).mean()
#fid_std = np.array(fids).std()
with open(savename_score, 'w+') as f:
f.write(str(fid_score))
#with open(savename_std, 'w+') as f:
# f.write(str(fid_std))
print('\nFID SCORE: {:.2f}'.format(fid_score))
return {"scalars": {"fid": fid_score}}
if __name__ == "__main__":
from edflow.debug import DebugDataset
from edflow.data.dataset import ProcessedDataset
D1 = DebugDataset(size=100)
D2 = DebugDataset(size=100)
P = lambda *args, **kwargs: {'image': np.ones([256, 256, 3])}
D1 = ProcessedDataset(D1, P)
D2 = ProcessedDataset(D2, P)
print(D1[0])
fid('.', D1, D2, {})
def test_seq_offset_fid_strat_unknown():
D = DebugDataset(size=12, offset=3)
with pytest.raises(ValueError):
S = SequenceDataset(D, 3, fid_key="label1", strategy="recolorize")
def __init__(self):
self.data = DebugDataset(size=10)
self.append_labels = True
def __init__(self):
self.data = DebugDataset(size=10)
# return {'pid': int(pid), 'vid': 0, 'fid': int(fid), 'action': act}
# D = DataFolder(r,
# rfn,
# lfn,
# ['pid', 'vid', 'fid'])
# for i in range(10):
# d = D[i]
# print(',\n '.join(['{}: {}'.format(k, v if not hasattr(v, 'shape')
# else v.shape)
# for k, v in d.items()]))
from edflow.debug import DebugDataset
D = DebugDataset()
def labels(data, i):
return {"fid": i}
D = ExtraLabelsDataset(D, labels)
print("D")
for k, v in D.labels.items():
print(k)
print(np.shape(v))
S = SequenceDataset(D, 2)
print("S")
for k, v in S.labels.items():
print(k)
print(np.shape(v))