Jeff Epler
196104084b
Mostly by providing a "numpy shim" for CircuitPython, try to make the numpy tests run on all three systems. (a "scipy shim" might also be useful?) However, there are test failures. Is it worth working through them and getting this to a point where it could be included?
51 lines
1.4 KiB
Python
51 lines
1.4 KiB
Python
try:
|
|
import numpy as np
|
|
except:
|
|
import ulab as np
|
|
|
|
def itemsize(x):
|
|
x = x.itemsize
|
|
if not isinstance(x, int): return x()
|
|
return x
|
|
|
|
def size(x):
|
|
x = x.size
|
|
if not isinstance(x, int): return x()
|
|
return x
|
|
|
|
def dtype(x):
|
|
x = x.dtype
|
|
print("dtype", x, type(x))
|
|
if not isinstance(x, (int, type, np.dtype)): return x()
|
|
return x
|
|
|
|
def shape(x):
|
|
x = x.shape
|
|
if callable(x): return x()
|
|
return x
|
|
|
|
a = np.array([1, 2, 3, 4], dtype=np.int8)
|
|
b = a.copy()
|
|
print(b)
|
|
a = np.array([[1,2,3],[4,5,6],[7,8,9]], dtype=np.int16)
|
|
b = a.copy()
|
|
print(b)
|
|
a = np.array([[1,2,3],[4,5,6],[7,8,9]], dtype=np.float)
|
|
b = a.copy()
|
|
print(b)
|
|
print(dtype(a))
|
|
print(a.flatten())
|
|
print(itemsize(np.array([1,2,3], dtype=np.uint8)))
|
|
print(itemsize(np.array([1,2,3], dtype=np.uint16)))
|
|
print(itemsize(np.array([1,2,3], dtype=np.int8)))
|
|
print(itemsize(np.array([1,2,3], dtype=np.int16)))
|
|
print(itemsize(np.array([1,2,3], dtype=np.float)))
|
|
print(shape(np.array([1,2,3], dtype=np.float)))
|
|
print(shape(np.array([[1],[2],[3]], dtype=np.float)))
|
|
print(np.array([[1],[2],[3]], dtype=np.float).reshape((1,3)))
|
|
print(size(np.array([[1],[2],[3]], dtype=np.float)))
|
|
print(size(np.array([1,2,3], dtype=np.float)))
|
|
print(np.array([1,2,3], dtype=np.uint8).tobytes())
|
|
print(np.array([1,2,3], dtype=np.int8).tobytes())
|
|
print(shape(np.array([1,2,3], dtype=np.float).transpose()))
|
|
print(shape(np.array([[1],[2],[3]], dtype=np.float).transpose()))
|