Help with linear optimisation problem

Hi folks,

I'm looking to conduct a linear optimisation and was hoping for some help on the matter.

I have 56 columns of data, with 256 rows each; lets call this set A. I have another column of data, with 56 rows; lets call this set B. I essentially want to correlate each of the 256 values in each column of set A to a single value in set B; lets call this set C.

I want to know which of the 256 values across set C give the best correlation between sets A and B. In other words, which row in set A gives the best correlation to the corresponding value in set B, across all columns of set A.

Any help on this matter would be most appreciated!

Answer 1

[Matt J]

Diffs=abs(reshape(B,1,1,[])-A);

d=min(Diffs,[],3);

[Arows,Acols]=find(d==min(d(:)));

[Teshan Rezel]

@Matt J Hi, thanks for this! I'm just having an issue with the first line of your code as it says reshape doesn't work on tables...any ideas on how I can get around this please?

[Matt J]

You can use table2array to convert them to ordinary matrices.

[Teshan Rezel]

@Matt J Thank you! I tried that but now it throws out this error, although my data is in tables, not cells!

Operator '-' is not supported for operands of type 'cell'.

[Matt J]

My solution assumes all variables are numeric matrices. If this isnt' the case for one/all of your variables, you need to convert them appropriately.

[Teshan Rezel]

@Matt J I've made sure that all values in the tables are numeric, but it now says :

Error using tabular/reshape (line 216)
Undefined function 'reshape' for input arguments of type 'table'.

I'm not sure what's causing this now!

[Matt J]

I've made sure that all values in the tables are numeric

Again, they shouldn't be tables. They should be matrices. To convert tables to matrices, see table2array() which I mentioned earlier.

[Teshan Rezel]

@Matt J thanks for this. I'm really not sure what I'm doing wrong as I've tried running the code snippet you sent me. I've attached the table that I think might be causing the issues. May I please ask you to have a look and tell me what I'm doing wrong? Apologies for this!

[Matt J]

What about A?

[Teshan Rezel]

@Matt J apologies, I've attached it as well!

[Matt J]

A=load('inputVariables').sum_Coke;
B=load('inputVariables').CRI;

B=B.CRI;

whos A B

  Name        Size             Bytes  Class     Attributes

  A         256x54            110592  double              
  B          54x1                432  double              

Diffs=abs(reshape(B,1,1,[])-A);

d=min(Diffs,[],3);

[Arows,Acols]=find(d==min(d(:)));

whos Arows Acols

  Name        Size            Bytes  Class     Attributes

  Acols      84x1               672  double              
  Arows      84x1               672  double              

[Teshan Rezel]

@Matt J outstanding, thank you!