Deleting zeros and NaN in a matrix

I have a matrix containing zerros and NaN's. Now i want to delete the rows if it consists zero or Nan

for ex if i have a matrix

A=[2  3  36  9
Nan  54  20  23
85  69  10  30
20  Nan  20  30
1  0  8  20
2  6  8  9]

i want to delete rows containing zeros and Nan's

so i will have output as

out=[2  3  36  9
85  69  10  30
2  6  8  9]

Please help,i have matrix containing 1078x8 values

Answer 1

[Image Analyst]

I'd do this:

nanRows = any(isnan(A), 2)
zeroRows = any(A==0, 2)
badRows = nanRows | zeroRows
A(badRows, :) = []

Of course you could compress all that into a single line if you want but I just used separate lines for tutorial purposes so you can see what's going on. In the command window you see:

A =
     2     3    36     9
   NaN    54    20    23
    85    69    10    30
    20   NaN    20    30
     1     0     8    20
     2     6     8     9
nanRows =
     0
     1
     0
     1
     0
     0
zeroRows =
     0
     0
     0
     0
     1
     0
badRows =
     0
     1
     0
     1
     1
     0
A =
     2     3    36     9
    85    69    10    30
     2     6     8     9

[Pat]

thanks Analyst

    Thanks Taylor but if i read values from excel sheet i get error 

    B = 

        [  2]    [56]    [NaN]    [ 5]    'a'
        [NaN]    [ 2]    [  0]    [20]    'b'
        [ 10]    [ 2]    [  3]    [ 6]    'c'
        [  1]    [20]    [ 30]    [40]    'd'

    Undefined function or method 'isnan' for input arguments of type 'cell'

please help

[Image Analyst]

Use cell2mat to extract out the first 4 columns to a numerical array. Right now it's a cell array.

[Pat]

i tried using ru comment using cell2mat for first 4 columns,i got

r

     2    56   NaN     5
     NaN     2     0    20
      10     2     3     6
       1    20    30    40

ten i tried concanateing r and last column i get error as 

Error using ==> horzcat
CAT arguments dimensions are not consistent.

please help i need output for above matrix as 

   10    2     3     6    'c'
   1    20     30    40    'd'

please help

[Image Analyst]

Why do you want to concatenate the last column when you just went to the trouble to strip it off? You don't need it to do what you asked originally.

If A is still in the workbook, you can do this first before you run my code above:

[A text raw]=xlsread('zz.xls');

to get A out of the workbook and into a variable in MATLAB.

Answer 2

[Kye Taylor]

Try

 A(any(isnan(A),2)|any(A==0,2),:) = []

[Pat]

Thanks Taylor but if i read values from excel sheet i get error 

B = 

      [  2]    [56]    [NaN]    [ 5]    'a'
      [NaN]    [ 2]    [  0]    [20]    'b'
      [ 10]    [ 2]    [  3]    [ 6]    'c'
      [  1]    [20]    [ 30]    [40]    'd'

Undefined function or method 'isnan' for input arguments of type 'cell'

please help

[Walter Roberson]

Only if you use the "raw" array, the third one returned by xlsread(). If you use the first array returned by xlsread() then it will be a numeric array rather than a cell array.

[Pat]

Walter i tried 

[num text raw]=xlsread('zz.xls');
 A=raw;
 nanRows = any(isnan(A), 2)
zeroRows = any(A==0, 2)
badRows = nanRows | zeroRows
A(badRows, :) = []

But i get error 

??? Undefined function or method 'isnan' for input arguments of type 'cell'.

[Walter Roberson]

[num text raw]=xlsread('zz.xls');
A=num;
nanRows = any(isnan(A), 2)
zeroRows = any(A==0, 2)
badRows = nanRows | zeroRows
A(badRows, :) = []