.-
help for ^mstore^
.-

Store and Retrieve Matrices with Dataset ----------------------------------------

^mstore aliasname ,^ [ ^f^rom^(^matrix expression^) m^ake^(^newmatrixname^)^ ^del^ete ^rep^lace]

syntax for getting list of stored matrices:

^mstore list^

Description -----------

^mstore^ is a utility to store matrices with a Stata dataset that can be easily retrieved in future sessions. The program uses character- istics so that the stored matrices are saved with the data. Full row and column names are maintained. Each matrix is stored using an alias name so that it can be assigned from matrix expressions and retrieved into any named matrix.

Options for ^mstore^ -------------------- ^aliasname^ is required and is used to name the stored copy of the matrix. This name is used by ^mstore^ to keep track of the stored matrices and can be any arbitrary text string without spaces.

^from(^matrixexpression^)^ identifies the matrix to be stored and is required if you wish to store a matrix. ^from^ will evaluate any argument as a matrix expression, so you could specify a matrix name, a matrix returned from an estimation command (e.g., ^from(e(b))^ ) or even an expression (e.g., ^from(syminv(A))^ ).

^make(^newmatrixname^)^ is used to specify the name to assign to a matrix you are retrieving. If ^make^ is not specified, then ^mstore^ will name the matrix using its aliasname (if you wish to use this syntax, then the aliasname must be a valid name for a Stata matrix). In either case, ^mstore^ will overwrite any existing matrix with the same n > ame.

^delete^ and ^replace^ are used to delete or replace the stored matrix aliasnam > e.

Examples --------

. ^regress mpg weight^

storing matrices . ^mstore myb, from(e(b))^ . ^mstore myv, from(e(V))^ . ^save myauto^

In another session, after using myauto.. list matrices: . ^mstore list^ 2 matrices found: myb myv

. ^mstore myb^ (creates matrix myb equal to stored matrix myb) . ^mstore myv, make(v2)^ (creates matrix v2 equal to stored matrix myv)

. ^mstore myv, delete^

Author ------ Michael Blasnik, Blasnik & Associates, Boston, MA mblasnik@@110.net