How do I count the number of NaNs in a Vector?

I have an nxm array a(n,m) with an unknown number of NaNs inside. I want to look for the column a(i,:) which contains the least number of Nans, how do I do this?

thank you for your help

Answer 1

[Daniel Shub]

This is a straight forward question and given you haven't shown any attempt to solve it, I am hesitant to provide an answer ...

Start by making some nonsense data

n = 10;
m = 5;
a = randn(n, m);
a(rand(numel(a), 1) < 0.25) = nan;

a =

    0.5377   -1.3499    0.6715    0.8884       NaN
    1.8339    3.0349   -1.2075   -1.1471       NaN
   -2.2588    0.7254    0.7172   -1.0689    0.3192
       NaN   -0.0631       NaN   -0.8095    0.3129
       NaN    0.7147    0.4889   -2.9443   -0.8649
   -1.3077   -0.2050    1.0347    1.4384       NaN
   -0.4336   -0.1241       NaN    0.3252   -0.1649
    0.3426       NaN   -0.3034   -0.7549    0.6277
    3.5784       NaN    0.2939    1.3703       NaN
       NaN    1.4172   -0.7873   -1.7115    1.1093

The ISNAN function tells you if an element is a nan (1) or not (0).

isnan(a)

ans =

     0     0     0     0     1
     0     0     0     0     1
     0     0     0     0     0
     1     0     1     0     0
     1     0     0     0     0
     0     0     0     0     1
     0     0     1     0     0
     0     1     0     0     0
     0     1     0     0     1
     1     0     0     0     0

Now we want to sum up all the ones in each column. We have the SUM function that does that

sum(isnan(a))

ans =

     3     2     2     0     4

To get the final answer we want to find which column has the smallest sum. The MIN function usually returns the smallest values, but if you read the documentation, the second output argument is the index of the minimum value.

[~, col] = min(sum(isnan(a)))

col =

     4

where the ~ notation allows for the first output argument to be ignored.

[Jason Baxter]

sorry, I haven't attempted it because I had no idea how to answer it, i'm not very good at Matlab yet. Thank you Daniel for your answer, I will answer more questions than I ask to keep the community going when I have time. Thank you for your answer! Jason

Answer 2

[Oleg Komarov]

[minval,idx] = min(sum(isnan(a)))

Answer 3

[Azzi Abdelmalek]

s=sum(isnan(a)); %
[dd,ff]=min(s);
find(s==dd) % gives the rows that contain the least "nan"

% it's possible there are many rows containing the least nan