long/wide - reshape confusion

[davidparsley]

I have a data set that looks like (see below), where x1-x4 are observations for 4 different individuals. I want to stack these into a panel, including a date column, and a column with the specific individual identifier (in this case simply x1, x2, x3, x4). I know how to do this with "stack date x1 date x2 date x3 date x4" (but I don't know how to "simply" get the individual identifiers). Now, my "real" question is: suppose I have 20,000 individuals (with not nice identifiers, e.g., slkijs1, sliiily, 889098, 1234145....). Clearly, my stack example above would be exceedingly tedious. I end up having to do this a lot, so any suggestions would be MOST appreciated. I'm sure I've overlooked some very basic information, but... thanks.

date	x1	x2	x3	x4
8-Jan-08	44.267	71.24366	7.064299	53.01147
9-Jan-08	21.462	12.1015	46.61863	20.77969
10-Jan-08	75.883	35.21949	2.601562	57.79749
11-Jan-08	83.377	66.84127	25.44249	90.14868
14-Jan-08	67.533	17.24804	67.00962	11.29083
15-Jan-08	67.819	68.5128	14.26676	76.55122
16-Jan-08	15.461	15.63645	95.98777	75.82653
17-Jan-08	19.519	40.68843	19.62687	86.06653
18-Jan-08	87.716	63.43431	81.71249	90.69047
21-Jan-08	16.507	64.96755	40.14056	0.718889

[Clyde Schechter]

You should be using the -reshape- command for this.

reshape long x, i(date) j(person_number)

If the real variable names are not a simple pattern like x1, x2, ... etc. then you should rename them so they are. So, if every variable except date corresponds to data for an individual, you could do something like this:

ds date, not
local individuals `r(varlist)'

rename (`individuals') x_=
reshape long x_, i(date) j(person) string

-reshape- is one of the most basic data management commands in Stata. You would be well advised to acquaint yourself with it by reading the corresponding chapter in the [D] volume of the PDF manuals that come with your Stata installation. (Select PDF Documentation on the Help menu.) It takes a bit of practice to get used to figuring out what goes in i() and what goes in j(), but after enough tries, it will suddenly "click" in your brain and you will just automatically know what to do.

[Jesse Wursten]

The example can be easily reshaped with the aptly named "reshape" command. The not-so-nice identifiers might be more tricky, but "reshape ..., ... string" might get you there.

[Nick Cox]

Clyde and Jesse rightly point to the string option of reshape.

As the identifiers are already part of variable names, there won't be a problem. They are going to end up in a string variable, and no rules will bite. You will need to map that string variable to a numeric variable to tsset or xtset, but in turn see e.g.

https://www.stata.com/support/faqs/d...p-identifiers/

[davidparsley]

You should be using the -reshape- command for this.

reshape long x, i(date) j(person_number)

If the real variable names are not a simple pattern like x1, x2, ... etc. then you should rename them so they are. So, if every variable except date corresponds to data for an individual, you could do something like this:

ds date, not
local individuals `r(varlist)'

rename (`individuals') x_=
reshape long x_, i(date) j(person) string

-reshape- is one of the most basic data management commands in Stata. You would be well advised to acquaint yourself with it by reading the corresponding chapter in the [D] volume of the PDF manuals that come with your Stata installation. (Select PDF Documentation on the Help menu.) It takes a bit of practice to get used to figuring out what goes in i() and what goes in j(), but after enough tries, it will suddenly "click" in your brain and you will just automatically know what to do.

Thanks, the -ds- command, which put all of my thousands of variable names into a neat variable list was the key, and I was unfamiliar with -ds-. On a side note, I haven't had the -reshape- epiphany yet despite repeated 'acquaintances' with the manuals. I find that most help examples are limited only to the very basic.

[Nick Cox]

Several suggestions can be found with search reshape. Here's a selection:

Book    . . . . . An Introduction to Stata for Health Researchers, 4th Edition
        . . . . . . . . . . . . . . . . . . . Svend Juul and Morten Frydenberg
        http://www.stata.com/bookstore/introduction-stata-health-researchers/

Book    . . . . . . . . . .  Data Management Using Stata: A Practical Handbook
        . . . . . . . . . . . . . . . . . . . . . . . . .  Michael N. Mitchell
        http://www.stata.com/bookstore/data-management-using-stata/

Video   . Data management: How to reshape data from long format to wide format
        4/17    http://www.youtube.com/watch?v=gkcYpw8CtCw
                This video demonstrates how to reshape data from long
                format to wide format.

Video   . Data management: How to reshape data from wide format to long format
        4/17    http://www.youtube.com/watch?v=Bx9kVdkr9oY
                This video demonstrates how to reshape data from wide
                format to long format.

FAQ     . . . . . . . . . . . . . . . . . . . . . . . .  Problems with reshape
        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
        2/14    I am having problems with the reshape command. Can
                you give further guidance?
                http://www.stata.com/support/faqs/data-management/
                problems-with-reshape/

Example . . . . . . . . . Stata learning module: Reshaping data:  long to wide
        . . . . . . . . . . . . . . . . . .  UCLA Academic Technology Services
        7/08    https://stats.idre.ucla.edu/stata/modules/reshaping-data-long-
                to-wide/

Example . . . . . . . . . Stata learning module:  Reshaping data: wide to long
        . . . . . . . . . . . . . . . . . .  UCLA Academic Technology Services
        7/08    https://stats.idre.ucla.edu/stata/modules/reshaping-data-wide-
                to-long/