Error combining splitapply and regress

Hello,

I have a table named 'T' containing a set of stock returns for different stocks as well as the market return for all days the stock returns are presented. I have divided my table into groups based on the stock and now I am trying to create a regression that uses the stock return as the dependant variable and a constant and the market return as the independant/explainatory variables, however I get the error message 'Invalid expression. When calling a function or indexing a variable, use parentheses. Otherwise, check for mismatched delimiters.'

I used the code as follow:

T.Groups = findgroups(T.Numero)
T.Ones = ones(length(T.MarketReturn),1);
X = splitapply(@regress(T.StockReturn,[T.Ones T.MarketReturn]),T,T.Groups);

Answer 1

[dpb]

First argument is just a function handle or an anonymous function -- you mixed metaphors by trying to call a function and call that a handle...

X = splitapply(@(x1,x2,y) regress(y,[x1 x2]),T.Ones,T.MarketReturn,T.StockReturn,T.Groups);

[Nabil Benhadda]

Thank you a lot ! However now I get the following error message:

Error using splitapply (line 132)

The function '@(x1,x2,y)regress(y,[x1,x2])' returned a non-scalar value when applied to the 1st group of data.

To compute nonscalar values for each group, create an anonymous function to return each value in a scalar cell:

@(x1,x2,y){regress(y,[x1,x2])}

Error in BeginWork (line 20)

X = splitapply(@(x1,x2,y) regress(y,[x1 x2]),Work.Ones,Work.MarketReturn,Work.StockReturn,Work.Groups);

[dpb]

Oh, yeah. splitapply doesn't automagically use a cell array when need.

Did you try the advice given in the error message? It shows exactly how to write for purpose; simply add the {} to return cell.

[Nabil Benhadda]

Thank you again !