def test_lookup_nan(self, writable):
# GH#21688 ensure we can deal with readonly memory views
xs = np.array([2.718, 3.14, np.nan, -7, 5, 2, 3])
xs.setflags(write=writable)
m = ht.Float64HashTable()
m.map_locations(xs)
tm.assert_numpy_array_equal(m.lookup(xs), np.arange(len(xs), dtype=np.intp))
def unique1d(values):
"""
Hash table-based unique
"""
if np.issubdtype(values.dtype, np.floating):
table = htable.Float64HashTable(len(values))
uniques = np.array(table.unique(_ensure_float64(values)),
dtype=np.float64)
elif np.issubdtype(values.dtype, np.datetime64):
table = htable.Int64HashTable(len(values))
uniques = table.unique(_ensure_int64(values))
uniques = uniques.view('M8[ns]')
elif np.issubdtype(values.dtype, np.timedelta64):
table = htable.Int64HashTable(len(values))
uniques = table.unique(_ensure_int64(values))
uniques = uniques.view('m8[ns]')
elif np.issubdtype(values.dtype, np.signedinteger):
table = htable.Int64HashTable(len(values))
uniques = table.unique(_ensure_int64(values))
elif np.issubdtype(values.dtype, np.unsignedinteger):
table = htable.UInt64HashTable(len(values))
uniques = table.unique(_ensure_uint64(values))
else:
# its cheaper to use a String Hash Table than Object
if lib.infer_dtype(values) in ['string']:
table = htable.StringHashTable(len(values))
else:
table = htable.PyObjectHashTable(len(values))
uniques = table.unique(_ensure_object(values))
return uniques
def test_add_signed_zeros(self):
# GH#21866 inconsistent hash-function for float64
# default hash-function would lead to different hash-buckets
# for 0.0 and -0.0 if there are more than 2^30 hash-buckets
# but this would mean 16GB
N = 4 # 12 * 10**8 would trigger the error, if you have enough memory
m = ht.Float64HashTable(N)
m.set_item(0.0, 0)
m.set_item(-0.0, 0)
assert len(m) == 1 # 0.0 and -0.0 are equivalent
def test_add_different_nans(self):
# GH#21866 inconsistent hash-function for float64
# create different nans from bit-patterns:
NAN1 = struct.unpack("d", struct.pack("=Q", 0x7FF8000000000000))[0]
NAN2 = struct.unpack("d", struct.pack("=Q", 0x7FF8000000000001))[0]
assert NAN1 != NAN1
assert NAN2 != NAN2
# default hash function would lead to different hash-buckets
# for NAN1 and NAN2 even if there are only 4 buckets:
m = ht.Float64HashTable()
m.set_item(NAN1, 0)
m.set_item(NAN2, 0)
assert len(m) == 1 # NAN1 and NAN2 are equivalent