Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, is already up and running.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: RE: Re: Extracting specific elements from a matrix

From   "Martin Weiss" <>
To   <>
Subject   st: RE: Re: Extracting specific elements from a matrix
Date   Mon, 5 Jul 2010 23:16:31 +0200


Vectors are not really part of Stata parlance, I am afraid, which may well
be a shock for those who previously used other products. 

When you say -generate-, you create a new column of data. If you want only
the first ten rows to be filled with data, as for the "y" variable in the
example, you can do that, but the rest (11 to _N) must contain something,
they cannot just be blank, as in some spreadsheet software products.

set obs 15
gen byte x=_n
gen byte y=_n in 1/10

You may also want to look at the -el()- function, btw.


-----Original Message-----
[] On Behalf Of Dani Tilley
Sent: Montag, 5. Juli 2010 23:10
Subject: st: Re: Extracting specific elements from a matrix

Much appreciated Martin. Incidentally, how do you go about creating an empty
vector in Stata? I noticed that -generate- creates a new variable of length
N,but what if I want a variable of length 10? I tried macro and local but
they appear to hold strings rather than floats. 

----- Original Message ----


Either talk to the coeff as "_b" (-help _variables-) or define a 
and extract from it:

sysuse auto, clear
reg price weight length rep78

di _b[weight]

mat A=e(b)
di A[1,1]

From: Dani Tilley <>
Sent: Mon, July 5, 2010 4:36:25 PM
Subject: Extracting specific elements from a matrix


I need to extract the first element of a matrix of coefficients and store it
on a variable, and repeat the process i times. I can set up the for loop and
do the storing, but I don't know how to access element [i,j] of a matrix A.

In particular, my matrix is e(b), the vector of coefficients from a
regression with k explanatory variables. I tried -e(b)[1,1]- but that didn't
take me far.

Any help is appreciated.

DF Tilley

*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2016 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index