erasing all the rows that begin with NaN

 I have a cell matric.

The first two columns of this matrix are

           [1x28 char]    [          NaN]

    [             NaN]    [          NaN]

    'MART'              'TOTATRIA'

    'PREGORY'             'TOSTE'   

    [             NaN]    [          NaN]

My goal is to erase these rows which have as a first element NaN

In the above example I want to erase row2 and the last one

thanks

Answer 1

[Andrei Bobrov]

just

out = A(~cellfun(@(x)all(isnan(x)),A(:,1)),:)

[Sabbas]

it works!thnx

Answer 2

[Thomas]

If A is your input cell matrix

out=cellfun(@isnan,A(:,1),'UniformOutput',false)
count=cellfun(@(x)any(x==1),out);
A(~count,:)

[Sabbas]

Hi. For some reason it does not work. The rows that begin with NaN still remain

thanks

[Thomas]

I'm using your old dataset. What error r u getting? or is it just not working

A = {[         NaN]    [     NaN]
[         NaN]    [     NaN]
'England'         [     NaN]
[         NaN]    [     NaN]
' AUSTRIA'        '  SMOKES'
'tetete'          'NIKE'
'jhjh khkhkh'     'GOODYS'
' BAND 1'         'HIT'
' BAND 2 KA 1'    'eerrrr'
' BAND 2 KA 2'    'dddddd'
' BAND 2 KA 3'    [     NaN]
' BAND 2 KA 4'    'YES' };

out=cellfun(@isnan,A(:,1),'UniformOutput',false)
count=cellfun(@(x)any(x==1),out);
A(~count,:)

Using this works for me.. Removes all rows beginning with NaN's

ans = 
    'England'         [     NaN]
    ' AUSTRIA'        '  SMOKES'
    'tetete'          'NIKE'    
    'jhjh khkhkh'     'GOODYS'  
    ' BAND 1'         'HIT'     
    ' BAND 2 KA 1'    'eerrrr'  
    ' BAND 2 KA 2'    'dddddd'  
    ' BAND 2 KA 3'    [     NaN]
    ' BAND 2 KA 4'    'YES'

[Sabbas]

I do not get any error. It is just not working

[Sabbas]

My matrix also contain numbers. For example, [45.66]

[Thomas]

Using the data you provided in the question

B={ 'Hello'    [          NaN]
[             NaN]    [          NaN]
'MART'              'TOTATRIA'
'PREGORY'             'TOSTE'
[             NaN]    [          NaN]}

A=B;
out=cellfun(@isnan,A(:,1),'UniformOutput',false)
count=cellfun(@(x)any(x==1),out);
A(~count,:)

ans = 
    'Hello'      [     NaN]
    'MART'       'TOTATRIA'
    'PREGORY'    'TOSTE'

[Thomas]

Should work irrespective of having numbers or not..

A={ 'Hello'    [          NaN]
[             NaN]    [          NaN]
[46.65] 'MORE'
'NEW' [46.65]
'MART'              'TOTATRIA'
'PREGORY'             'TOSTE'
[             NaN]    [          NaN]};
out=cellfun(@isnan,A(:,1),'UniformOutput',false)
count=cellfun(@(x)any(x==1),out);
A(~count,:)

[Jan]

@Binish: Is there a difference between "any(x==1)" and "any(x)"?

[Thomas]

Gives the same answer in this case with both any(x==1) and any(x) since both are working on output of isnan...

[Sabbas]

it works for me when I choose the first two columns. But not when I use the whole matrix

[Sabbas]

I used

count=cellfun(@(x)any(x==1),out);

A(count==1,:)=[]

and it seems to work

could you plese verify that the code is equivalent to the one you provided?

thanks

Answer 3

[Jan]

A(cellfun(@(x) isequalwithequalnans(x, NaN), C), :)