Clever Geek Handbook

Using numpy output in boolean form

I'm using Python 2.7.5 on Windows 7. For some reason python doesn't like it when I use one of the sizes of my numpy array with a comparator in an if statement:

a = np.array([1,2,3,4])
# reshapes so that array has two dimensions
if len(np.shape(a)) == 1:
    a = np.reshape(a, (1, np.shape(a)))

b = np.shape(a)[0]

if b <= 3:
    print 'ok'

I am creating a 1D numpy array (actually "a" is an input, which can be 1D or 2D). Then I reformat this to form a 2D numpy array. I am trying to use the size of a newly created dimension as a comparator and I am getting an error: "TypeError: Integer required"

I also tried "int (b)" to convert a long integer to a simple integer in an if statement and it gives the same error. If I do "type (b)" it gives me "the type is" long ". It seems to me that I have done this before without any problem, but I can’t find any examples. It’s something with the way I change 1D -array to 2D array? Any help is appreciated.

+2


source to share


3 answers


Problematic line a = np.reshape(a, (1, np.shape(a)))

.

To add an axis to the origin a

, I would suggest using:



a = a[np.newaxis, ...]

print a.shape # (1, 4)

or None

does the same as np.newaxis

.

+1


source


You are creating a tuple with np.shape, so you are passing (1,(4,))

, so the error has nothing to do with your if, this is what is happening inside the if you need to use np.shape(a)[0]

, but I'm not really sure what you are trying to do:

 np.shape(a)[0]


Or simply a.shape[0]

+3


source


It looks like you are trying to do the same as np.atleast_2d

:

def atleast_2d(*arys):   # *arys handles multiple arrays
    res = []
    for ary in arys:
        ary = asanyarray(ary)
        if len(ary.shape) == 0 :
            result = ary.reshape(1, 1)
        elif len(ary.shape) == 1 :  # looks like your code!
            result = ary[newaxis,:]
        else :
            result = ary
        res.append(result)
    if len(res) == 1:
        return res[0]
    else:
        return res

In [955]: a=np.array([1,2,3,4])
In [956]: np.atleast_2d(a)
Out[956]: array([[1, 2, 3, 4]])

or this is a list:

In [961]: np.atleast_2d([1,2,3,4])
Out[961]: array([[1, 2, 3, 4]])

you can also check the attribute ndim

:a.ndim==1

0


source






More articles:


All Articles