Searching a cell array of percentile scores based on age and gender

I have one table in matlab and two cell arrays. The table is a list of ID numbers, age, sex, and test scores. The second two cell arrays are percentile rankings, one table for males and the other for females. What I am trying to do first is use the sex to select the correct percentile rankings array. Then, using the age, find the correct column. The ages are located in row 1. Then within that column find the matching score. Once the matching score is found the percentile ranking is in the first column for that row.

This is an example of the Subject table

      PartID     Age     Sex    Test
      _______    ____    ___    __________

      'Y01'     19      1      ''        
      'Y02'     22      1      '58.39'    
      'Y03'     21      1      '61.28'    
      'Y04'     NaN    NaN     ''        
      'Y05'     19      1      '63.55'

This is an example of the male percentile table. Ages are in Row 1 and percentile rankings are Column 1. The inside of the array is the test scores.

      'Males'    [     19]    [     20]    [     21]    [     22]
      [   97]    [63.5500]    [     63]    [62.4500]    [61.9000]
      [   96]    [62.3700]    [61.8300]    [61.2800]    [60.7400]
      [   95]    [61.1800]    [60.6500]    [60.1100]    [59.5700]
      [   94]    [59.9800]    [59.4500]    [58.9200]    [58.3900]

For subject Y02 it would determine they had a percentile of 94. Y03 would be 96. Y05 would be 97. Then any subject with a missing value would be skipped.

Any suggestions on how to accomplish this?

Thank you

Answer 1

[Andrei Bobrov]

S = table(arrayfun(@(x)sprintf('Y%02d',x),(1:5)','un',0),[19;22;21;nan;19],...
             [1;1;1;nan;1],{'','58.39','61.28','','63.55'},'VariableNames',...
                                             {'PartID' 'Age' 'Sex','Test'});
M = {      'Males'    [     19]    [     20]    [     21]    [     22]
      [   97]    [63.5500]    [     63]    [62.4500]    [61.9000]
      [   96]    [62.3700]    [61.8300]    [61.2800]    [60.7400]
      [   95]    [61.1800]    [60.6500]    [60.1100]    [59.5700]
      [   94]    [59.9800]    [59.4500]    [58.9200]    [58.3900]};

[ii,jj] = ndgrid([M{2:end,1}],[M{1,2:end}]);
M2 = [M{2:end,2:end}]';
mm = [jj(:),ones(numel(ii),1),M2(:),ii(:)];
mm1 = [S{:,{'Age','Sex'}},str2double(S({:,{'Test'}})];
[a,b,c] = intersect(mm1,mm(:,1:3),'rows');
out = sortrows([S.PartID(b),num2cell(mm(c,4))],1);

[Daniel]

Thank you for the response.

I had some trouble with the S = table portion of the script and I added a ' to the end of the curly braces and that seemed to fix it

S = table(arrayfun(@(x)sprintf('Y%02d',x),(1:5)','un',0),[19;22;21;nan;19],...
             [1;1;1;nan;1],{'','58.39','61.28','','63.55'}','VariableNames',...
                                             {'PartID' 'Age' 'Sex','Test'});

But I am having an issue with this portion

mm1 = [S{:,{'Age','Sex'}},str2double(S(:,{'Test'}))];

I get this error "Error using horzcat Dimensions of matrices being concatenated are not consistent."

[Andrei Bobrov]

corrected

[Daniel]

Thank you very much for the help

Answer 2

[Katarzyna Wieciorek]

I would try find

[row,col] = find(X==score(id)) and extract percentile from this row in column 1 if col corresponds to age

for skipping you need if else