使用numpy random时，选择多个重复值吗？

Question

I would like to know how can I select the values with more than one repetition when using numpy random to randomly generate numbers. I am doing the following:

1) First generate the numbers between 1-10

ran = ceil( np.random.random(10)*10 )
print ran
[  2.   9.   8.   9.  10.   8.  10.   7.   1.   1.]

2) Now using itemfreq from scipy.stats to get the frequency table where the second column is the frequency.

freq_tmp = itemfreq(ran)
print freq_tmp
[[  1.   2.]
[  2.   1.]
[  7.   1.]
[  8.   2.]
[  9.   2.]
[ 10.   2.]]

print freq_tmp[0:][:,1] 

[ 2.  1.  1.  2.  2.  2.]

this show me only the frequency but I would like to know the values with more than 1 repetition, in this example should be:

[1. 8. 9. 10]

Thank you for your help!

Answer 1

You are almost there. You can just put a > in the slicer like so:

import numpy as np

a = np.array([[1,2],[2,1],[7,1],[8,2],[9,2],[10,2]])

print(a[a[:,1] > 1][:,0])

output:

[ 1  8  9 10]

Answer 2

Another numpy solution :

In [57]: ran=randint(1,11,10)

In [58]: ran
Out[58]: array([3, 4, 6, 1, 9, 4, 2, 8, 6, 8])

In [59]: uniqs,cnts=np.unique(ran,return_counts=True)

In [60]: uniqs,cnts
Out[60]: (array([1, 2, 3, 4, 6, 8, 9]), array([1, 1, 1, 2, 2, 2, 1], dtype=int64))

In [61]: uniqs[cnts>=2]
Out[61]: array([4, 6, 8])