error in selecting data

i have a data of 2 columns

d=   [12] 'A'
     [23] 'B' 
     [18] 'H'

d1=[23 45]

out = [d1, d(ismember(d(:,1),d1(:,1),'rows'),1:end)]

if i do this i get error

Input must be cell arrays of strings

please help

Answer 1

[Sven]

Hi FIR,

The error comes because the ismember() function expects either a matrix of numbers or a cell array of strings. Since d is a cell, you provided a cell array of numbers which is unfortunately invalid. Try converting to numbers like the following:

dMask = ismember(cell2mat(d(:,1)), d1);

Secondly, you were trying to concat d1 (which is a 1x2 matrix) with various rows of a cell (ie, an nx2 cell):

out = [d1, d(dMask,:)]

This won't work on two counts:

1. Cells won't concat with matrices - one of them needs to be converted.

2. d1 is 1x2, which won't concat if your result has more than 1 rows.

 ------

So, I've updated my answer as below from the comments you made. Firstly, we have your inputs d and d1 as follows:

d = {12 'A'; 23 'B'; 18 'H'; 12 'P'};
d1 = [23 45;18 10];

And what you actually want to do is concat d1 with its corresponding indices into d. Note that this means you're asking which entries in the first column of d1 are members of the first column of d... Note that this is the opposite ordering from your original question, but it seems more logical. Try the following two lines, I think that they answer your question:

[dMask, indices] = ismember(d1(:,1), cell2mat(d(:,1)));
[num2cell(d1(dMask,:)) d(indices,2)];

Does this help?

[FIR]

sven please suggest .assuming that d1 has more rows

for example d has 4000 rows and 62 column,d has 10 rows

[FIR]

Sven i get an error wen executing ur command

Error using ==> cell2mat

Too many input arguments.

Error in ==> qq at 8

dMask = ismember(cell2mat(d(:,1), d1));

[Sven]

So answer this:

A = [1 2];

B = [10 20; 15 40];

Here "A" has 1 row, "B" has 2 rows. Please clarify what your final answer would look like if you combined A and B.

[Sven]

Oh, sorry, there was a typo: I missed a ")". It's fixed now

[FIR]

but sven ur A and B data is wrong

from my dat d and d1

i need output as

23 45 B

if

data is

d= [12] 'A'

[23] 'B'

[18] 'H'

[12] 'P'

d1=[23 45;18 10]

but sven ur A and B data is wrong

from my dat d and d1

i need output as

23 45 B

if

data is

d= [12] 'A'

[23] 'B'

[18] 'H'

[12] 'P'

d1=[23 45;18 10]

i NEED OUTPUT AS

23 45 B

18 10 H

[Sven]

Answer updated. What do you think?

[FIR]

THANKS A LOT Sven,this is what i was expecting

[FIR]

Another one question I have 4000 rows and 5 columns I have to divide this data into 5 groups so that i will have 800 rows and 25 columns,plz help

[Sven]

If the question's answered, hit accept :)

Does the following do what you wanted?

MAT_4000by5 = rand(4000,5);

MAT_800by25 = reshape(MAT_4000by5,800,25);

[FIR]

Thanks Sven i have 2 more questions ,

I have 5 columns and 800 rows and the last row contains percentage ranging from (90.20 to 99.98),first i want to round off that percentage ans have to display it withs all its column and ten i want to select two percentage which is highest and want to dispaly with all its 5 columns

output must be

1 3 1 0 99

2 3 1 0 98

next i want to find five fold crodd validation error for the third column,by using in built function,please help

[Sven]

I'm a bit lost on what you want to do here, and it seems quite a different topic to the original question. I suggest hitting "accept" to the original question, and then asking a new question specifically about "how to format rows columns for display". If you can make a short example showing exactly the input you have and the output you want, I'm sure it will get a clear answer.

[FIR]

Sven u have not answered to my question