st: RE: RE: Price Index

["A.Silva" <as454@kent.ac.uk>]

Nick, 
I see. Thanks very much,
Andres
________________________________________
From: owner-statalist@hsphsun2.harvard.edu [owner-statalist@hsphsun2.harvard.edu] On Behalf Of Nick Cox [n.j.cox@durham.ac.uk]
Sent: 05 April 2011 18:26
To: 'statalist@hsphsun2.harvard.edu'
Subject: st: RE: Price Index

Let me outline solutions for two situations.

1. Suppose you have a single item whose price you are monitoring.

Then

su price if time == 495

gen priceindex = 100 * price / r(mean)

I am clearly assuming that you want 100 as your base.

2. However, I imagine you would not be asking if that was the problem. Your question implies something like a panel structure.

Then you could reach for something more like

egen priceindex = total(price / (time == 495)), by(panelid)

or

egen priceindex = total(price * (time == 495)), by(panelid)

followed by

replace priceindex = 100 * price / priceindex

The logic is that -time- of 495 occurs just once in each panel. You should check that.

You could squeeze that into one command, no doubt, but I guess with some loss of clarity.

Nick
n.j.cox@durham.ac.uk

A.Silva

I need to create a price index using a base price.
How can I indicate to Stata to use always the same observation as a base (in my example time==495)?
For example, my variables are:
time price
495 20
496 22
497 23


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/