Append values unless in last column

Question

I would like to add values to an array if they occur after a 100. This works fine if the column which contains a 100 isn't the last column (as the next columns value in that row will be appended to an array). But an IndexError occurs if the 100 value occurs in the last column as there's no lastcol + 1 value.

col1 col2 col3
nan  100  60
100  95   98
nan  nan  100

Now:

values = [60,95,IndexError]; 

Ideal:

values = [60,95,100];

My code:

x1 = np.where(#table == 100.0)[0]; x2 = np.where(#table == 100.0)[1]

# np.where returns a tuple containing the (x,y) locations of   
the 100 values in the table. e.g. [(0,1,2),(1,0,2)] for the  
table in the above example.

for i,j in zip(x1,x2):
   values.append(out[i][j+1]);
# Above attempts to add values

EDIT

col1 col2 col3 col4
nan  100  60   50
100  95   98   70
nan  nan  100  80
nan  nan  nan  100
nan  nan  100  100

Desired: Get values after 100 occurs in row and append it to the 'values' array. Also, note that nan will occur in each row before a 100 values occurs.

values = [60,95,80,100,100];

The values above occur in each row after 100 (order is important).

Answer 1

I'm not sure what you expect exactly, because I don't understand how you can get values = [60,95,80,100,100];

But I guess you were close to the solution

Actually expected value !

Rule : We want the first value after the first occurrence of 100, and if first value of 100 is the last one, then take it.

In [1]: A
Out[1]: 
array([[  nan,  100.,   60.,   50.],
       [ 100.,   95.,   98.,   70.],
       [  nan,   nan,  100.,   80.],
       [  nan,   nan,   nan,  100.],
       [  nan,   nan,  100.,  100.],
       [ 100.,    4.,  100.,    5.]])


In [2]: B = np.where(A==100)

As A is 2D :

• B[0] is lines position of 100 in A
• B[1] is columns position of 100 in A

Then :

In [3]: value = []
In [4]: for j in set(B[0]): # set returns unique values in array B[0]
        idx = A[j].tolist().index(100) # I get the index of the first occurrence of 100  in line j of matrix A (list.index(value,[start,[stop]]) gives position of first occurrence from start to stop in list )
        if idx+1 >= len(A[j]): # if it's last column ... 
            value.append(A[(j,idx)]) # ... add 100
        else:
            value.append(A[(j,idx+1)]) # if not, add value after 100

 In [5]: value
 Out[5]: [60.0, 95.0, 80.0, 100.0, 100.0, 4.0]

Expected value 1

with ternary operator

If your expected value is [ 60., 95., 80., 100., 100., 100.] (Take first value after 100 or take 100 if 100 is last)

Then :

In [1]: A
Out[1]: 
array([[  nan,  100.,   60.,   50.],
       [ 100.,   95.,   98.,   70.],
       [  nan,   nan,  100.,   80.],
       [  nan,   nan,   nan,  100.],
       [  nan,   nan,  100.,  100.]])

In [2]: B = np.where(A==100)

In [3]: A[(B[0],[b+1 if b+1<len(A[0]) else b for b in B[1]])]
Out[3]: array([  60.,   95.,   80.,  100.,  100.,  100.])

Expected value 2

If your expected value is [ 60., 95., 80., 100.], (take first value after 100 and if 100 is last take nothing) then :

value = []
for i,j in zip(B[0],B[1]):
    if j+1<len(A[0]):
         value.append(A[i,j+1])

value being now [60.0, 95.0, 80.0, 100.0]

Tell me if you don't understand something.