From e1e319f844edb2e6a22aa815ca42806d47c6cf5f Mon Sep 17 00:00:00 2001 From: Rory Yorke Date: Sun, 23 Jun 2019 14:46:23 +0200 Subject: [PATCH] Use numpy.longdouble instead of numpy.float128 in testing numpy.float128 doesn't exist on all platforms, e.g., Windows 64-bit. When numpy.float128 does exist (e.g., Linux 64-bit), it's typically an alias for numpy.longdouble. --- control/tests/xferfcn_input_test.py | 28 ++++++++++++++-------------- 1 file changed, 14 insertions(+), 14 deletions(-) diff --git a/control/tests/xferfcn_input_test.py b/control/tests/xferfcn_input_test.py index 0471e885..0d6ca56f 100644 --- a/control/tests/xferfcn_input_test.py +++ b/control/tests/xferfcn_input_test.py @@ -7,7 +7,7 @@ import numpy as np from numpy import int, int8, int16, int32, int64 -from numpy import float, float16, float32, float64, float128 +from numpy import float, float16, float32, float64, longdouble from numpy import all, ndarray, array from control.xferfcn import _clean_part @@ -73,7 +73,7 @@ def test_clean_part_tuple(self): def test_clean_part_all_scalar_types(self): """Test single scalar value for all valid data types.""" - for dtype in [int, int8, int16, int32, int64, float, float16, float32, float64, float128]: + for dtype in [int, int8, int16, int32, int64, float, float16, float32, float64, longdouble]: num = dtype(1) num_ = _clean_part(num) @@ -92,7 +92,7 @@ def test_clean_part_np_array(self): def test_clean_part_all_np_array_types(self): """Test scalar value in numpy array of ndim=0 for all data types.""" - for dtype in [int, int8, int16, int32, int64, float, float16, float32, float64, float128]: + for dtype in [int, int8, int16, int32, int64, float, float16, float32, float64, longdouble]: num = np.array(1, dtype=dtype) num_ = _clean_part(num) @@ -102,7 +102,7 @@ def test_clean_part_all_np_array_types(self): def test_clean_part_all_np_array_types2(self): """Test numpy array for all types.""" - for dtype in [int, int8, int16, int32, int64, float, float16, float32, float64, float128]: + for dtype in [int, int8, int16, int32, int64, float, float16, float32, float64, longdouble]: num = np.array([1, 2], dtype=dtype) num_ = _clean_part(num) @@ -112,7 +112,7 @@ def test_clean_part_all_np_array_types2(self): def test_clean_part_list_all_types(self): """Test list of a single value for all data types.""" - for dtype in [int, int8, int16, int32, int64, float, float16, float32, float64, float128]: + for dtype in [int, int8, int16, int32, int64, float, float16, float32, float64, longdouble]: num = [dtype(1)] num_ = _clean_part(num) assert isinstance(num_, list) @@ -121,7 +121,7 @@ def test_clean_part_list_all_types(self): def test_clean_part_list_all_types2(self): """List of list of numbers of all data types.""" - for dtype in [int, int8, int16, int32, int64, float, float16, float32, float64, float128]: + for dtype in [int, int8, int16, int32, int64, float, float16, float32, float64, longdouble]: num = [dtype(1), dtype(2)] num_ = _clean_part(num) assert isinstance(num_, list) @@ -130,7 +130,7 @@ def test_clean_part_list_all_types2(self): def test_clean_part_tuple_all_types(self): """Test tuple of a single value for all data types.""" - for dtype in [int, int8, int16, int32, int64, float, float16, float32, float64, float128]: + for dtype in [int, int8, int16, int32, int64, float, float16, float32, float64, longdouble]: num = (dtype(1),) num_ = _clean_part(num) assert isinstance(num_, list) @@ -139,7 +139,7 @@ def test_clean_part_tuple_all_types(self): def test_clean_part_tuple_all_types2(self): """Test tuple of a single value for all data types.""" - for dtype in [int, int8, int16, int32, int64, float, float16, float32, float64, float128]: + for dtype in [int, int8, int16, int32, int64, float, float16, float32, float64, longdouble]: num = (dtype(1), dtype(2)) num_ = _clean_part(num) assert isinstance(num_, list) @@ -184,7 +184,7 @@ def test_clean_part_list_list_list_floats(self): def test_clean_part_list_list_array(self): """List of list of numpy arrays for all valid types.""" - for dtype in int, int8, int16, int32, int64, float, float16, float32, float64, float128: + for dtype in int, int8, int16, int32, int64, float, float16, float32, float64, longdouble: num = [[array([1, 1], dtype=dtype), array([2, 2], dtype=dtype)]] num_ = _clean_part(num) @@ -195,7 +195,7 @@ def test_clean_part_list_list_array(self): def test_clean_part_tuple_list_array(self): """Tuple of list of numpy arrays for all valid types.""" - for dtype in int, int8, int16, int32, int64, float, float16, float32, float64, float128: + for dtype in int, int8, int16, int32, int64, float, float16, float32, float64, longdouble: num = ([array([1, 1], dtype=dtype), array([2, 2], dtype=dtype)],) num_ = _clean_part(num) @@ -206,7 +206,7 @@ def test_clean_part_tuple_list_array(self): def test_clean_part_list_tuple_array(self): """List of tuple of numpy array for all valid types.""" - for dtype in int, int8, int16, int32, int64, float, float16, float32, float64, float128: + for dtype in int, int8, int16, int32, int64, float, float16, float32, float64, longdouble: num = [(array([1, 1], dtype=dtype), array([2, 2], dtype=dtype))] num_ = _clean_part(num) @@ -217,7 +217,7 @@ def test_clean_part_list_tuple_array(self): def test_clean_part_tuple_tuples_arrays(self): """Tuple of tuples of numpy arrays for all valid types.""" - for dtype in int, int8, int16, int32, int64, float, float16, float32, float64, float128: + for dtype in int, int8, int16, int32, int64, float, float16, float32, float64, longdouble: num = ((array([1, 1], dtype=dtype), array([2, 2], dtype=dtype)), (array([3, 4], dtype=dtype), array([4, 4], dtype=dtype))) num_ = _clean_part(num) @@ -229,7 +229,7 @@ def test_clean_part_tuple_tuples_arrays(self): def test_clean_part_list_tuples_arrays(self): """List of tuples of numpy arrays for all valid types.""" - for dtype in int, int8, int16, int32, int64, float, float16, float32, float64, float128: + for dtype in int, int8, int16, int32, int64, float, float16, float32, float64, longdouble: num = [(array([1, 1], dtype=dtype), array([2, 2], dtype=dtype)), (array([3, 4], dtype=dtype), array([4, 4], dtype=dtype))] num_ = _clean_part(num) @@ -241,7 +241,7 @@ def test_clean_part_list_tuples_arrays(self): def test_clean_part_list_list_arrays(self): """List of list of numpy arrays for all valid types.""" - for dtype in int, int8, int16, int32, int64, float, float16, float32, float64, float128: + for dtype in int, int8, int16, int32, int64, float, float16, float32, float64, longdouble: num = [[array([1, 1], dtype=dtype), array([2, 2], dtype=dtype)], [array([3, 3], dtype=dtype), array([4, 4], dtype=dtype)]] num_ = _clean_part(num)