How can an equation be obtained using four series of data?

  ET  Age  BMI  RI 
 
  5/78824  88/7762  19/878  0/889028 
 
  6/00752  70/5228  23/9439  0/888978 
 
  6/28914  47/2447  25/3443  0/910558 
 
  6/4053  72/5024  15/6769  0/823959 
 
  6/45901  85/3978  26/278  0/862904 
 
  6/57075  72/5024  15/6769  0/932137 
 
  6/75893  90/019  28/4013  0/847686 
 
  7/13733  85/301  26/3233  0/821629 
 
  7/33723  77/0434  26/9558  0/860541 
 
  7/38556  84/3978  25/3296  0/910309 
 
  7/62961  80/454  24/7475  0/860475 
 
  7/78118  72/9315  26/1427  0/849619 
 
  8/36594  84/7977  24/6069  0/849486 
 
  8/7508  72/9315  26/0074  0/810442 
 
  8/81851  39/1794  24/8329  0/821248 
 
  8/83789  53/336  18/5686  0/782286 
 
  8/86374  74/7087  19/7733  0/730337 
 
  9/14642  64/4662  27/8748  0/749752 
 
  9/32599  71/7813  27/8748  0/682617 
 
  9/39358  64/239  30/1637  0/840596 
 
  9/43234  58/7996  30/1637  0/762673 
 
  9/78382  52/4174  30/1638  0/790729 
 
  9/89245  60/8628  30/1638  0/719282 
 
  9/90645  39/2019  23/0179  0/691143 
 
  9/92691  60/4927  30/1638  0/650017 
 
  9/94509  58/7996  30/1637  0/760392 
 
  9/95908  60/8628  30/1638  0/732253 
 
  9/97093  60/8628  30/1637  0/708443 
 
  10/403  39/2019  23/1918  0/721331 
 
  10/9092  51/4455  19/638  0/87921 
 
  10/9298  53/2766  22/1918  0/690911 
 
  10/934  52/699  23/7041  0/829426 
 
  10/9415  52/699  23/7041  0/814274 
 
  11/0438  52/699  23/7041  0/608642 
 
  11/0663  73/5338  23/7041  0/710359 
 
  11/091  39/2091  23/0179  0/660574 
 
  11/9059  59/8242  23/9402  0/638747 
 
  11/924  54/9399  23/9301  0/749122 
 
  11/9477  61/227  23/9402  0/701502 
 
  11/9648  65/4585  23/9363  0/814042 
 
  11/9671  58/8343  28/3574  0/66254 
 
  11/9736  58/8342  23/9402  0/649553 
 
  11/9821  51/006  23/9363  0/779409 
 
  11/9841  51/006  23/9301  0/922253 
 
  12/0025  54/9399  23/9301  0/738283 
 
  12/0154  61/227  23/9301  0/712308 
 
  12/0348  58/8343  28/3574  0/673346 
 
  12/0627  54/9399  23/9301  0/764241 
 
  12/9838  53/2766  22/0179  0/822468 
 
  13/0205  45/2211  22/0179  0/748874 
 
  13/0635  53/2867  22/3793  0/662292 
 
  13/07  53/2867  22/3793  0/649304 
 
  13/0731  51/9933  20/1184  0/789983 
 
  13/1582  53/2767  22/0179  0/618984 
 
  13/1722  59/6786  22/3793  0/590845 
 
  14/0159  65/4585  23/4337  0/657747 

[Rik]

What kind of function would you like to fit to this data? What is your goal?

[ebrahim yzdnmnsh]

I want to get an equation to get ET using the three other data sets.

3 set together

[ebrahim yzdnmnsh]

its mean for example :

ET=age(x)+or-MRI*(y)+or-RI*(z)

[John D'Errico]

Are those slashes meant as decimal points?

[ebrahim yzdnmnsh]

yes

[ebrahim yzdnmnsh]

I want to get like this equation:

exaplme=

ET = -9.356+(0.084*RI)+(0.069*AGE)+(0.02*BMI)

It's not important to get exactly these numbers, just a rational equation to get those numbers.

[ebrahim yzdnmnsh]

Even if a few numbers are obtained correctly, it's okay. In the whole, it does not want to get all the exact numbers. An equation is enough.

[Walter Roberson]

 A=[ ones(numel(RI), 1), RI(:), AGE(:), BMI(:) ];
 b = ET(:);
  x = A\b;
  ETpredict = @(RI, AGE, BMI) x(1) + x(2)*RI + x(3)*AGE + x(4)*BMI;

This will give you a prediction function along the lines you indicated. It will not do a good job.

Answer 1

[Sulaymon Eshkabilov]

Hi,

Here is the way how to create equations from your data:

ABCD=[5/78824  88/7762  19/878  0/889028;
   6/00752  70/5228  23/9439  0/888978 ;
   6/28914  47/2447  25/3443  0/910558 ;
   6/4053  72/5024  15/6769  0/823959 ;
   6/45901  85/3978  26/278  0/862904 ;
   6/57075  72/5024  15/6769  0/932137; 
   6/75893  90/019  28/4013  0/847686 ;
   7/13733  85/301  26/3233  0/821629 ;
   7/33723  77/0434  26/9558  0/860541; 
   7/38556  84/3978  25/3296  0/910309 ;
   7/62961  80/454  24/7475  0/860475 ;
   7/78118  72/9315  26/1427  0/849619; 
   8/36594  84/7977  24/6069  0/849486 ;
   8/7508  72/9315  26/0074  0/810442 ;
   8/81851  39/1794  24/8329  0/821248; 
   8/83789  53/336  18/5686  0/782286 ;
   8/86374  74/7087  19/7733  0/730337; 
   9/14642  64/4662  27/8748  0/749752 ;
   9/32599  71/7813  27/8748  0/682617 ;
   9/39358  64/239  30/1637  0/840596 ;
   9/43234  58/7996  30/1637  0/762673; 
   9/78382  52/4174  30/1638  0/790729 ;
   9/89245  60/8628  30/1638  0/719282 ;
   9/90645  39/2019  23/0179  0/691143 ;
   9/92691  60/4927  30/1638  0/650017 ;
   9/94509  58/7996  30/1637  0/760392 ;
   9/95908  60/8628  30/1638  0/732253 ;
   9/97093  60/8628  30/1637  0/708443 ;
   10/403  39/2019  23/1918  0/721331 ;
   10/9092  51/4455  19/638  0/87921 ;
   10/9298  53/2766  22/1918  0/690911; 
   10/934  52/699  23/7041  0/829426 ;
   10/9415  52/699  23/7041  0/814274; 
   11/0438  52/699  23/7041  0/608642 ;
   11/0663  73/5338  23/7041  0/710359; 
   11/091  39/2091  23/0179  0/660574 ;
   11/9059  59/8242  23/9402  0/638747; 
   11/924  54/9399  23/9301  0/749122 ;
   11/9477  61/227  23/9402  0/701502 ;
   11/9648  65/4585  23/9363  0/814042; 
   11/9671  58/8343  28/3574  0/66254 ;
   11/9736  58/8342  23/9402  0/649553; 
   11/9821  51/006  23/9363  0/779409 ;
   11/9841  51/006  23/9301  0/922253 ;
   12/0025  54/9399  23/9301  0/738283; 
   12/0154  61/227  23/9301  0/712308 ;
   12/0348  58/8343  28/3574  0/673346; 
   12/0627  54/9399  23/9301  0/764241 ;
   12/9838  53/2766  22/0179  0/822468 ;
   13/0205  45/2211  22/0179  0/748874 ;
   13/0635  53/2867  22/3793  0/662292 ;
   13/07  53/2867  22/3793  0/649304 ;
   13/0731  51/9933  20/1184  0/789983; 
   13/1582  53/2767  22/0179  0/618984 ;
   13/1722  59/6786  22/3793  0/590845 ;
   14/0159  65/4585  23/4337  0/657747] ;

syms x y z
EQN = ABCD(:,1) == ABCD(:,2)*x+ABCD(:,3)*y+ABCD(:,4)*z;
%% ABCD(:,1) = ET; ABCD(:,2) = Age; ABCD(:,2) = BMI; ABCD(:,2) = RI; 

[ebrahim yzdnmnsh]

Hello

Thank you

But I meant to get an equation for ET to get ET with the data of other rows.

That is, when we put the data of the other columns in the equation, we get the ET value.

Answer 2

[Walter Roberson]

"That is, when we put the data of the other columns in the equation, we get the ET value."

You are not likely going to be able to do that.

Plot ET on the X axis and Age on the Y axes. Draw a line across at about Age 53. Notice how many times it intercepts the plot. Any entry of age 53 could be referring to any of those locations.

Likewise plot with BMI on the Y axes and draw a line across at 22 and see how many times you intercept the plot. Any entry of 22 could be referring to any of those locations.

Likewise with RI on the Y axes and draw a line across about 0.71 and see how many times you intercept the plot. Any entry of 0.71 could be referring to any of those locations.

In order to be able to predict ET from the other columns, your thesis would have to be that, "Yes it is true that there are lots of duplicate values, but the combination is always unique, that there is no place that is a duplicate on all three values, and therefore by looking carefully at the combination, you can figure out which ET it derived from." I do not think you would be able to sustain that thesis, though.