This is my current code:
import numpy as np
vec0 = [1, 2, 3]
vec1 = [2, 3, 4]
vec2 = [3, 4, 5]
vec3 = [4, 5, 6]
for k in range(0, 4):
globals()['mean%s' % k] = np.mean('vec'+str(k))
I'm getting this error:
TypeError: cannot perform reduce with flexible type
I want to this result.
mean0 = np.mean(vec0)
mean1 = np.mean(vec1)
mean2 = np.mean(vec2)
mean3 = np.mean(vec3)
It is an anti-pattern to call variables by name. In case you need to perform a task on multiple objects, you can construct a collection (a tuple, a list, etc.) of these objects. For instance:
all_vecs = [vec0, vec1, vec2, vec3]
Furthermore you can now easily process the averages in bulk with numpy by specifying an axis parameter:
all_means = np.mean(all_vecs,axis=1)
then:
>>> all_means
array([ 2., 3., 4., 5.])
You can do this:
vecs = [vec0, vec1, vec2, vec3]
and then :
globals()['mean%s' % k] = np.mean(vecs[k])
Although, you should avoid this approach and maintain a 2D Array instead, which will be easier for computing means using the row-axis.
What I mean is, instead of this:
vec0 = [1, 2, 3]
vec1 = [2, 3, 4]
vec2 = [3, 4, 5]
vec3 = [4, 5, 6]
you can have something like this :
vecs = np.array([[1, 2, 3], [2, 3, 4], [3, 4, 5], [4, 5, 6]])
and then you can simply calculate the means like this:
meansArray = vecs.mean(axis=1)
which will have your mean0, mean1, mean2, mean3 at the respective indices.
You can do it by changing:
globals()['mean%s' % k] = np.mean('vec'+str(k))
to
globals()['mean%s' % k] = np.mean(globals()['vec%s' % k])
But I highly recommend using a vector instead:
all_vec = [vec0, vec1, vec2, vec3]
