Fixed #227 Warn users that inputs must be float

seanlaw · seanlaw · commit 231ccb1dd210 · 2020-08-30T23:18:15.000-04:00
diff --git a/stumpy/aamp.py b/stumpy/aamp.py
@@ -279,9 +279,6 @@ def aamp(T_A, m, T_B=None, ignore_trivial=True):
     if T_B.ndim != 1:  # pragma: no cover
         raise ValueError(f"T_B is {T_B.ndim}-dimensional and must be 1-dimensional. ")
 
-    core.check_dtype(T_A)
-    core.check_dtype(T_B)
-
     core.check_window_size(m)
 
     if ignore_trivial is False and core.are_arrays_equal(T_A, T_B):  # pragma: no cover
diff --git a/stumpy/aamped.py b/stumpy/aamped.py
@@ -69,9 +69,6 @@ def aamped(dask_client, T_A, m, T_B=None, ignore_trivial=True):
     if T_B.ndim != 1:  # pragma: no cover
         raise ValueError(f"T_B is {T_B.ndim}-dimensional and must be 1-dimensional. ")
 
-    core.check_dtype(T_A)
-    core.check_dtype(T_B)
-
     core.check_window_size(m)
 
     if ignore_trivial is False and core.are_arrays_equal(T_A, T_B):  # pragma: no cover
diff --git a/stumpy/aampi.py b/stumpy/aampi.py
@@ -80,7 +80,9 @@ def __init__(self, T, m, excl_zone=None, egress=True):
             If set to `True`, the oldest data point in the time series is removed and
             the time series length remains constant rather than forever increasing
         """
-        self._T = np.asarray(T)
+        self._T = T.copy()
+        self._T = np.asarray(self._T)
+        core.check_dtype(self._T)
         self._m = m
         self._n = self._T.shape[0]
         if excl_zone is not None:  # pragma: no cover
diff --git a/stumpy/core.py b/stumpy/core.py
@@ -118,7 +118,8 @@ def check_dtype(a, dtype=np.floating):  # pragma: no cover
         If the array type does not match `dtype`
     """
     if not np.issubdtype(a.dtype, dtype):
-        msg = f"{dtype} type expected but found {a.dtype}"
+        msg = f"{dtype} type expected but found {a.dtype}\n"
+        msg += "Please change your input `dtype` with `.astype(float)`"
         raise TypeError(msg)
 
     return True
@@ -852,6 +853,7 @@ def preprocess(T, m):
     """
     T = T.copy()
     T = np.asarray(T)
+    check_dtype(T)
 
     T[np.isinf(T)] = np.nan
     M_T, Σ_T = compute_mean_std(T, m)
@@ -889,6 +891,7 @@ def preprocess_non_normalized(T, m):
     """
     T = T.copy()
     T = np.asarray(T)
+    check_dtype(T)
 
     T[np.isinf(T)] = np.nan
     T_subseq_isfinite = np.all(np.isfinite(rolling_window(T, m)), axis=1)
diff --git a/stumpy/gpu_aamp.py b/stumpy/gpu_aamp.py
@@ -437,9 +437,6 @@ def gpu_aamp(T_A, m, T_B=None, ignore_trivial=True, device_id=0):
             "For multidimensional STUMP use `stumpy.mstump` or `stumpy.mstumped`"
         )
 
-    core.check_dtype(T_A)
-    core.check_dtype(T_B)
-
     core.check_window_size(m)
 
     if ignore_trivial is False and core.are_arrays_equal(T_A, T_B):  # pragma: no cover
diff --git a/stumpy/gpu_stump.py b/stumpy/gpu_stump.py
@@ -461,9 +461,6 @@ def gpu_stump(T_A, m, T_B=None, ignore_trivial=True, device_id=0):
             "For multidimensional STUMP use `stumpy.mstump` or `stumpy.mstumped`"
         )
 
-    core.check_dtype(T_A)
-    core.check_dtype(T_B)
-
     core.check_window_size(m)
 
     if ignore_trivial is False and core.are_arrays_equal(T_A, T_B):  # pragma: no cover
diff --git a/stumpy/mstump.py b/stumpy/mstump.py
@@ -619,9 +619,6 @@ def mstump(T, m, include=None, discords=False):
         err = f"T is {T_A.ndim}-dimensional and must be at least 1-dimensional"
         raise ValueError(f"{err}")
 
-    core.check_dtype(T_A)
-    core.check_dtype(T_B)
-
     core.check_window_size(m)
 
     if include is not None:
diff --git a/stumpy/mstumped.py b/stumpy/mstumped.py
@@ -80,9 +80,6 @@ def mstumped(dask_client, T, m, include=None, discords=False):
         err = f"T is {T_A.ndim}-dimensional and must be at least 1-dimensional"
         raise ValueError(f"{err}")
 
-    core.check_dtype(T_A)
-    core.check_dtype(T_B)
-
     core.check_window_size(m)
 
     if include is not None:
diff --git a/stumpy/scrump.py b/stumpy/scrump.py
@@ -180,21 +180,21 @@ def prescrump(T_A, m, T_B=None, s=None):
 
     See Algorithm 2
     """
-    T_A = np.asarray(T_A)
     T_A = T_A.copy()
-    T_A[np.isinf(T_A)] = np.nan
+    T_A = np.asarray(T_A)
     core.check_dtype(T_A)
+    T_A[np.isinf(T_A)] = np.nan
 
     if T_B is None:
         T_B = T_A
         excl_zone = int(np.ceil(m / 4))
     else:
         excl_zone = None
 
-    T_B = np.asarray(T_B)
     T_B = T_B.copy()
-    T_B[np.isinf(T_B)] = np.nan
+    T_B = np.asarray(T_B)
     core.check_dtype(T_B)
+    T_B[np.isinf(T_B)] = np.nan
 
     core.check_window_size(m)
 
@@ -376,9 +376,6 @@ def __init__(
                 "For multidimensional STUMP use `stumpy.mstump` or `stumpy.mstumped`"
             )
 
-        core.check_dtype(self._T_A)
-        core.check_dtype(self._T_B)
-
         core.check_window_size(self._m)
 
         if self._ignore_trivial is False and core.are_arrays_equal(
diff --git a/stumpy/stamp.py b/stumpy/stamp.py
@@ -136,9 +136,6 @@ def stamp(T_A, T_B, m, ignore_trivial=False):
     if T_B.ndim != 1:  # pragma: no cover
         raise ValueError(f"T_B is {T_B.ndim}-dimensional and must be 1-dimensional. ")
 
-    core.check_dtype(T_A)
-    core.check_dtype(T_B)
-
     core.check_window_size(m)
     subseq_T_B = core.rolling_window(T_B, m)
     excl_zone = int(np.ceil(m / 2))
diff --git a/stumpy/stomp.py b/stumpy/stomp.py
@@ -85,9 +85,6 @@ def _stomp(T_A, m, T_B=None, ignore_trivial=True):
     if T_B.ndim != 1:  # pragma: no cover
         raise ValueError(f"T_B is {T_B.ndim}-dimensional and must be 1-dimensional. ")
 
-    core.check_dtype(T_A)
-    core.check_dtype(T_B)
-
     core.check_window_size(m)
 
     if ignore_trivial is False and core.are_arrays_equal(T_A, T_B):  # pragma: no cover
diff --git a/stumpy/stump.py b/stumpy/stump.py
@@ -499,9 +499,6 @@ def stump(T_A, m, T_B=None, ignore_trivial=True):
             "For multidimensional STUMP use `stumpy.mstump` or `stumpy.mstumped`"
         )
 
-    core.check_dtype(T_A)
-    core.check_dtype(T_B)
-
     core.check_window_size(m)
 
     if ignore_trivial is False and core.are_arrays_equal(T_A, T_B):  # pragma: no cover
diff --git a/stumpy/stumped.py b/stumpy/stumped.py
@@ -132,9 +132,6 @@ def stumped(dask_client, T_A, m, T_B=None, ignore_trivial=True):
             "For multidimensional STUMP use `stumpy.mstump` or `stumpy.mstumped`"
         )
 
-    core.check_dtype(T_A)
-    core.check_dtype(T_B)
-
     core.check_window_size(m)
 
     if ignore_trivial is False and core.are_arrays_equal(T_A, T_B):  # pragma: no cover
diff --git a/stumpy/stumpi.py b/stumpy/stumpi.py
@@ -79,7 +79,9 @@ def __init__(self, T, m, excl_zone=None, egress=True):
             If set to `True`, the oldest data point in the time series is removed and
             the time series length remains constant rather than forever increasing
         """
-        self._T = np.asarray(T)
+        self._T = T.copy()
+        self._T = np.asarray(self._T)
+        core.check_dtype(self._T)
         self._m = m
         self._n = self._T.shape[0]
         if excl_zone is not None:  # pragma: no cover
diff --git a/tests/test_aamp.py b/tests/test_aamp.py
@@ -20,6 +20,11 @@
 substitution_values = [np.nan, np.inf]
 
 
+def test_aamp_int_input():
+    with pytest.raises(TypeError):
+        aamp(np.arange(10), 5)
+
+
 @pytest.mark.parametrize("T_A, T_B", test_data)
 def test_aamp_self_join(T_A, T_B):
     m = 3
diff --git a/tests/test_aamped.py b/tests/test_aamped.py
@@ -29,12 +29,18 @@ def dask_cluster():
 window_size = [8, 16, 32]
 
 
+def test_aamp_int_input(dask_cluster):
+    with pytest.raises(TypeError):
+        with Client(dask_cluster) as dask_client:
+            aamped(dask_client, np.arange(10), 5)
+
+
 @pytest.mark.filterwarnings("ignore:numpy.dtype size changed")
 @pytest.mark.filterwarnings("ignore:numpy.ufunc size changed")
 @pytest.mark.filterwarnings("ignore:numpy.ndarray size changed")
 @pytest.mark.filterwarnings("ignore:\\s+Port 8787 is already in use:UserWarning")
 @pytest.mark.parametrize("T_A, T_B", test_data)
-def test_stumped_self_join(T_A, T_B, dask_cluster):
+def test_aamped_self_join(T_A, T_B, dask_cluster):
     with Client(dask_cluster) as dask_client:
         m = 3
         left = naive.aamp(T_B, m)
@@ -49,7 +55,7 @@ def test_stumped_self_join(T_A, T_B, dask_cluster):
 @pytest.mark.filterwarnings("ignore:numpy.ndarray size changed")
 @pytest.mark.filterwarnings("ignore:\\s+Port 8787 is already in use:UserWarning")
 @pytest.mark.parametrize("T_A, T_B", test_data)
-def test_stumped_self_join_df(T_A, T_B, dask_cluster):
+def test_aamped_self_join_df(T_A, T_B, dask_cluster):
     with Client(dask_cluster) as dask_client:
         m = 3
         left = naive.aamp(T_B, m)
@@ -65,7 +71,7 @@ def test_stumped_self_join_df(T_A, T_B, dask_cluster):
 @pytest.mark.filterwarnings("ignore:\\s+Port 8787 is already in use:UserWarning")
 @pytest.mark.parametrize("T_A, T_B", test_data)
 @pytest.mark.parametrize("m", window_size)
-def test_stump_self_join_larger_window(T_A, T_B, m, dask_cluster):
+def test_aamped_self_join_larger_window(T_A, T_B, m, dask_cluster):
     with Client(dask_cluster) as dask_client:
         if len(T_B) > m:
             left = naive.aamp(T_B, m)
@@ -82,7 +88,7 @@ def test_stump_self_join_larger_window(T_A, T_B, m, dask_cluster):
 @pytest.mark.filterwarnings("ignore:\\s+Port 8787 is already in use:UserWarning")
 @pytest.mark.parametrize("T_A, T_B", test_data)
 @pytest.mark.parametrize("m", window_size)
-def test_stump_self_join_larger_window_df(T_A, T_B, m, dask_cluster):
+def test_aamped_self_join_larger_window_df(T_A, T_B, m, dask_cluster):
     with Client(dask_cluster) as dask_client:
         if len(T_B) > m:
             left = naive.aamp(T_B, m)
@@ -98,7 +104,7 @@ def test_stump_self_join_larger_window_df(T_A, T_B, m, dask_cluster):
 @pytest.mark.filterwarnings("ignore:numpy.ndarray size changed")
 @pytest.mark.filterwarnings("ignore:\\s+Port 8787 is already in use:UserWarning")
 @pytest.mark.parametrize("T_A, T_B", test_data)
-def test_stumped_A_B_join(T_A, T_B, dask_cluster):
+def test_aamped_A_B_join(T_A, T_B, dask_cluster):
     with Client(dask_cluster) as dask_client:
         m = 3
         left = naive.aamp(T_A, m, T_B=T_B)
@@ -113,7 +119,7 @@ def test_stumped_A_B_join(T_A, T_B, dask_cluster):
 @pytest.mark.filterwarnings("ignore:numpy.ndarray size changed")
 @pytest.mark.filterwarnings("ignore:\\s+Port 8787 is already in use:UserWarning")
 @pytest.mark.parametrize("T_A, T_B", test_data)
-def test_stumped_A_B_join_df(T_A, T_B, dask_cluster):
+def test_aamped_A_B_join_df(T_A, T_B, dask_cluster):
     with Client(dask_cluster) as dask_client:
         m = 3
         left = naive.aamp(T_A, m, T_B=T_B)
diff --git a/tests/test_aampi.py b/tests/test_aampi.py
@@ -9,6 +9,11 @@
 substitution_values = [np.nan, np.inf]
 
 
+def test_aampi_int_input():
+    with pytest.raises(TypeError):
+        aampi(np.arange(10), 5)
+
+
 def test_aampi_self_join():
     m = 3
     zone = int(np.ceil(m / 4))
diff --git a/tests/test_gpu_aamp.py b/tests/test_gpu_aamp.py
@@ -30,6 +30,11 @@
 substitution_values = [np.nan, np.inf]
 
 
+def test_gpu_aamp_int_input():
+    with pytest.raises(TypeError):
+        gpu_aamp(np.arange(10), 5, ignore_trivial=True)
+
+
 @pytest.mark.parametrize("T_A, T_B", test_data)
 def test_gpu_aamp_self_join(T_A, T_B):
     m = 3
diff --git a/tests/test_gpu_stump.py b/tests/test_gpu_stump.py
@@ -30,6 +30,11 @@
 substitution_values = [np.nan, np.inf]
 
 
+def test_gpu_aamp_int_input():
+    with pytest.raises(TypeError):
+        gpu_stump(np.arange(10), 5, ignore_trivial=True)
+
+
 @pytest.mark.parametrize("T_A, T_B", test_data)
 def test_gpu_stump_self_join(T_A, T_B):
     m = 3
diff --git a/tests/test_mstump.py b/tests/test_mstump.py
@@ -144,6 +144,11 @@ def test_naive_mstump():
     npt.assert_almost_equal(left_I, right_I)
 
 
+def test_mstump_int_input():
+    with pytest.raises(TypeError):
+        mstump(np.arange(20).reshape(2, 10), 5)
+
+
 @pytest.mark.parametrize("T, m", test_data)
 def test_mstump(T, m):
     excl_zone = int(np.ceil(m / 4))
diff --git a/tests/test_mstumped.py b/tests/test_mstumped.py
@@ -21,6 +21,12 @@ def dask_cluster():
 ]
 
 
+def test_mstumped_int_input(dask_cluster):
+    with pytest.raises(TypeError):
+        with Client(dask_cluster) as dask_client:
+            mstumped(dask_client, np.arange(20).reshape(2, 10), 5)
+
+
 @pytest.mark.filterwarnings("ignore:\\s+Port 8787 is already in use:UserWarning")
 @pytest.mark.parametrize("T, m", test_data)
 def test_mstumped(T, m, dask_cluster):
diff --git a/tests/test_scrump.py b/tests/test_scrump.py
@@ -197,6 +197,11 @@ def test_prescrump_self_join_larger_window(T_A, T_B, m):
             npt.assert_almost_equal(left_I, right_I)
 
 
+def test_scrump_int_input():
+    with pytest.raises(TypeError):
+        scrump(np.arange(10), 5, ignore_trivial=True, percentage=1.0, pre_scrump=False)
+
+
 @pytest.mark.parametrize("T_A, T_B", test_data)
 @pytest.mark.parametrize("percentages", percentages)
 def test_scrump_self_join(T_A, T_B, percentages):
diff --git a/tests/test_stamp.py b/tests/test_stamp.py
@@ -49,6 +49,12 @@ def test_stamp_mass_PI(T_A, T_B):
     npt.assert_almost_equal(left_right_I, right_right_I)
 
 
+def test_stamp_int_input():
+    with pytest.raises(TypeError):
+        T = np.arange(10)
+        stamp(T, T, 5, ignore_trivial=True)
+
+
 @pytest.mark.parametrize("T_A, T_B", test_data)
 def test_stamp_self_join(T_A, T_B):
     m = 3
diff --git a/tests/test_stomp.py b/tests/test_stomp.py
@@ -21,6 +21,11 @@
 substitution_values = [np.nan, np.inf]
 
 
+def test_stomp_int_input():
+    with pytest.raises(TypeError):
+        stomp(np.arange(10), 5, ignore_trivial=True)
+
+
 @pytest.mark.parametrize("T_A, T_B", test_data)
 def test_stomp_self_join(T_A, T_B):
     m = 3
diff --git a/tests/test_stump.py b/tests/test_stump.py
@@ -21,6 +21,11 @@
 substitution_values = [np.nan, np.inf]
 
 
+def test_stump_int_input():
+    with pytest.raises(TypeError):
+        stump(np.arange(10), 5)
+
+
 @pytest.mark.parametrize("T_A, T_B", test_data)
 def test_stump_self_join(T_A, T_B):
     m = 3
diff --git a/tests/test_stumped.py b/tests/test_stumped.py
@@ -29,6 +29,12 @@ def dask_cluster():
 window_size = [8, 16, 32]
 
 
+def test_stumped_int_input(dask_cluster):
+    with pytest.raises(TypeError):
+        with Client(dask_cluster) as dask_client:
+            stumped(dask_client, np.arange(10), 5)
+
+
 @pytest.mark.filterwarnings("ignore:numpy.dtype size changed")
 @pytest.mark.filterwarnings("ignore:numpy.ufunc size changed")
 @pytest.mark.filterwarnings("ignore:numpy.ndarray size changed")
diff --git a/tests/test_stumpi.py b/tests/test_stumpi.py
@@ -9,6 +9,11 @@
 substitution_values = [np.nan, np.inf]
 
 
+def test_stumpi_int_input():
+    with pytest.raises(TypeError):
+        stumpi(np.arange(10), 5)
+
+
 def test_stumpi_self_join():
     m = 3
     zone = int(np.ceil(m / 4))

Original file line number	Diff line number	Diff line change
`@@ -437,9 +437,6 @@ def gpu_aamp(T_A, m, T_B=None, ignore_trivial=True, device_id=0):`
`437`	`437`	"For multidimensional STUMP use `stumpy.mstump` or `stumpy.mstumped`"
`438`	`438`	`)`
`439`	`439`
`440`		`- core.check_dtype(T_A)`
`441`		`- core.check_dtype(T_B)`
`442`		`-`
`443`	`440`	`core.check_window_size(m)`
`444`	`441`
`445`	`442`	`if ignore_trivial is False and core.are_arrays_equal(T_A, T_B): # pragma: no cover`
Original file line number	Diff line number	Diff line change
`@@ -461,9 +461,6 @@ def gpu_stump(T_A, m, T_B=None, ignore_trivial=True, device_id=0):`
`461`	`461`	"For multidimensional STUMP use `stumpy.mstump` or `stumpy.mstumped`"
`462`	`462`	`)`
`463`	`463`
`464`		`- core.check_dtype(T_A)`
`465`		`- core.check_dtype(T_B)`
`466`		`-`
`467`	`464`	`core.check_window_size(m)`
`468`	`465`
`469`	`466`	`if ignore_trivial is False and core.are_arrays_equal(T_A, T_B): # pragma: no cover`