I have a case of strange behaviour of numpy diff:
a = list(img_arr[y_coord_1,:])
print a
print np.diff(a)
>>[62, 62, 62, 62, 62, 62, 62, 62, 63, 62, 96, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 66, 63, 64, 64, 63, 63, 63, 63, 63, 63, 63, 63, 63, 63, 64, 65, 65, 64, 63, 63, 63, 63, 63, 63, 63, 64, 64, 63, 63, 63, 63, 63, 64, 65, 65, 64, 63, 63, 63, 63]
>>[ 0 0 0 0 0 0 0 1 255 34 2 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 224 253 1 0 255
0 0 0 0 0 0 0 0 0 1 1 0 255 255 0 0 0 0
0 0 1 0 255 0 0 0 0 1 1 0 255 255 0 0 0]
Now, when I run this in the terminal I get the correct answer of
array([ 0, 0, 0, 0, 0, 0, 0, 1, -1, 34, 2, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, -32, -3, 1,
0, -1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1,
0, -1, -1, 0, 0, 0, 0, 0, 0, 1, 0, -1, 0,
0, 0, 0, 1, 1, 0, -1, -1, 0, 0, 0])
What sort of things can cause this - I am using a few other compiled libraries in this script, if this of relevance
Edit: I've just spotted that its the negative numbers that are wrong - and the upper limit is very suspicious. Looks like a dtype issue..
.tolist()
is a better way of converting an array to a list (or nested lists). It carries the conversion all the way down. list()
just iterates on one level. And since an array is already iterable, I don't think list(anarray)
does anything useful.
start with an array:
In [789]: z
Out[789]: array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9], dtype=uint8)
In [790]: type(list(z)[0])
Out[790]: numpy.uint8
list()
is the same as this list comprehension:
In [791]: type([i for i in z][0])
Out[791]: numpy.uint8
the correct list conversion
In [792]: type(z.tolist()[0])
Out[792]: int
Why were you using list()
in the first place? You didn't need it for diff
. If overflow is an issue, dtype conversion is better.
np.diff
will turn a list back into an array before taking the differences.
In [793]: np.diff(z.tolist())
Out[793]: array([1, 1, 1, 1, 1, 1, 1, 1, 1])
In [794]: np.diff(list(z))
Out[794]: array([1, 1, 1, 1, 1, 1, 1, 1, 1], dtype=uint8)
In [795]: np.diff(z.astype('int'))
Out[795]: array([1, 1, 1, 1, 1, 1, 1, 1, 1])
In [796]: np.diff(z)
Out[796]: array([1, 1, 1, 1, 1, 1, 1, 1, 1], dtype=uint8)
In [797]: np.array(list(z))
Out[797]: array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9], dtype=uint8)
The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.