def test_ismember2d(self):
b = np.reshape([0, 0, 0, 1, 1, 1], [3, 2])
locb = np.array([0, 1, 0, 2, 2, 1])
lia = np.array([True, True, True, True, True, True, False, False])
a = np.r_[b[locb, :], np.array([[2, 1], [1, 2]])]
lia_, locb_ = bnum.ismember2d(a, b)
assert np.all(lia == lia_) & np.all(locb == locb_)
def test_uuids_intersections(self):
ntotal = 500
nsub = 17
nadd = 3
eids = uuid2np([uuid.uuid4() for _ in range(ntotal)])
np.random.seed(42)
isel = np.floor(np.argsort(np.random.random(nsub)) / nsub *
ntotal).astype(np.int16)
sids = np.r_[eids[isel, :],
uuid2np([uuid.uuid4() for _ in range(nadd)])]
np.random.shuffle(sids)
# check the intersection
v, i0, i1 = intersect2d(eids, sids)
assert np.all(eids[i0, :] == sids[i1, :])
assert np.all(np.sort(isel) == np.sort(i0))
v_, i0_, i1_ = np.intersect1d(eids[:, 0],
sids[:, 0],
return_indices=True)
assert np.setxor1d(v_, v[:, 0]).size == 0
assert np.setxor1d(i0, i0_).size == 0
assert np.setxor1d(i1, i1_).size == 0
for a, b in zip(ismember2d(sids, eids),
ismember(sids[:, 0], eids[:, 0])):
assert np.all(a == b)
# check conversion to numpy back and forth
uuids = [uuid.uuid4() for _ in np.arange(4)]
np_uuids = uuid2np(uuids)
assert np2uuid(np_uuids) == uuids
def test_ismember2d_uuids(self):
nb = 20
na = 500
np.random.seed(42)
a = np.random.randint(0, nb + 3, na)
b = np.arange(nb)
lia, locb = bnum.ismember(a, b)
bb = np.random.randint(low=np.iinfo(np.int64).min,
high=np.iinfo(np.int64).max,
size=(nb, 2),
dtype=np.int64)
aa = np.zeros((na, 2), dtype=np.int64)
aa[lia, :] = bb[locb, :]
lia_, locb_ = bnum.ismember2d(aa, bb)
assert np.all(lia == lia_) & np.all(locb == locb_)
bb[:, 0] = 0
aa[:, 0] = 0
# if the first column is equal, the distinction is to be made on the second\
assert np.unique(bb[:, 1]).size == nb
lia_, locb_ = bnum.ismember2d(aa, bb)
assert np.all(lia == lia_) & np.all(locb == locb_)
def _update_cache(self, ses, dataset_types):
"""
:param ses: session details dictionary as per Alyx response
:param dataset_types:
:return: is_updated (bool): if the cache was updated or not
"""
save = False
pqt_dsets = _ses2pandas(ses, dtypes=dataset_types)
# if the dataframe is empty, return
if pqt_dsets.size == 0:
return
# if the cache is empty create the cache variable
elif self._cache.size == 0:
self._cache = pqt_dsets
save = True
# the cache is not empty and there are datasets in the query
else:
isin, icache = ismember2d(pqt_dsets[['id_0', 'id_1']].to_numpy(),
self._cache[['id_0', 'id_1']].to_numpy())
# check if the hash / filesize fields have changed on patching
heq = (self._cache['hash'].iloc[icache].to_numpy() ==
pqt_dsets['hash'].iloc[isin].to_numpy())
feq = np.isclose(self._cache['file_size'].iloc[icache].to_numpy(),
pqt_dsets['file_size'].iloc[isin].to_numpy(),
rtol=0,
atol=0,
equal_nan=True)
eq = np.logical_and(heq, feq)
# update new hash / filesizes
if not np.all(eq):
self._cache.iloc[icache,
4:6] = pqt_dsets.iloc[np.where(isin)[0],
4:6].to_numpy()
save = True
# append datasets that haven't been found
if not np.all(isin):
self._cache = self._cache.append(
pqt_dsets.iloc[np.where(~isin)[0]])
self._cache = self._cache.reindex()
save = True
if save:
# before saving makes sure pandas did not cast uuids in float
typs = [
t for t, k in zip(self._cache.dtypes, self._cache.keys())
if 'id_' in k
]
assert (all(map(lambda t: t == np.int64, typs)))
# if this gets too big, look into saving only when destroying the ONE object
parquet.save(self._cache_file, self._cache)
def test_rf_map(self):
"""
"""
# Simulate fake rfmap data
test_frames = np.full((60, 15, 15), 128, dtype='uint8')
# Test on and off individually
test_frames[10:20, 8, 8] = 0
test_frames[25:35, 10, 13] = 255
# Test that interleaved are detected correctly
test_frames[40:50, 4, 9] = 0
test_frames[42:52, 6, 10] = 255
test_frames[42:55, 11, 4] = 0
test_frames[50:60, 8, 8] = 0
test_times = np.arange(60)
rf_map = {}
rf_map['times'] = test_times
rf_map['frames'] = test_frames
rf_map_times, rf_map_pos, rf_stim_frames = passive.get_on_off_times_and_positions(
rf_map)
assert (all(rf_map_times == test_times))
assert (rf_map_pos.shape == (15 * 15, 2))
assert (len(rf_stim_frames['on']) == 15 * 15)
assert (len(rf_stim_frames['off']) == 15 * 15)
# Off is for the 0 ones
assert (all(rf_stim_frames['off'][ismember2d(
rf_map_pos, np.array([[8, 8]]))[0]][0][0] == [10, 50]))
assert (rf_stim_frames['off'][ismember2d(rf_map_pos,
np.array([[4, 9]
]))[0]][0][0] == 40)
assert (rf_stim_frames['off'][ismember2d(rf_map_pos,
np.array([[11, 4]
]))[0]][0][0] == 42)
# On is for the 255 ones
assert (rf_stim_frames['on'][ismember2d(rf_map_pos,
np.array([[10, 13]
]))[0]][0][0] == 25)
assert (rf_stim_frames['on'][ismember2d(rf_map_pos,
np.array([[6, 10]
]))[0]][0][0] == 42)
# Next test that the firing rate function works
# Basically just make one square responsive
spike_times = np.arange(25, 35, 0.01)
spike_depths = 500 * np.ones_like(spike_times)
rf_map_avg, depths = passive.get_rf_map_over_depth(rf_map_times,
rf_map_pos,
rf_stim_frames,
spike_times,
spike_depths,
x_lim=[0, 60])
non_zero = np.where(rf_map_avg['on'] != 0)
assert (np.argmin(np.abs(depths - 500)) == non_zero[0][0])
assert (all(non_zero[1] == 10))
assert (all(non_zero[2] == 13))
assert (np.all(rf_map_avg['off'] == 0))
rf_svd = passive.get_svd_map(rf_map_avg)
# Make sure that the one responsive element is non-zero
assert (rf_svd['on'][non_zero[0][0]][non_zero[1][0], non_zero[2][0]] !=
0)
# But that all the rest are zero
rf_svd['on'][non_zero[0][0]][non_zero[1][0], non_zero[2][0]] = 0
assert (np.all(np.isclose(np.vstack(rf_svd['on']), 0)))
assert (np.all(np.vstack(rf_svd['off']) == 0))