*Open the log file; the path and name of the files are specific to your computer;

*change the directory to where you have saved the files for use in lab 1
cd "C:\Documents and Settings\Sandrah Eckel\Desktop\LDA lab1"

log using "lab1.log"
infile week1-week9 using "pigs.data"
save "pigs.dta", replace

*load a previously saved version of this data that already has an id variable
use "pigs.stata.dta", clear

*re-load the dataset we had just saved 
use "pigs.dta", clear

*Genearate the id=observation number
generate id=_n
move id week1
list
list in 1/5
describe 
save "pigs.dta", replace

*reshape the dataset into long format;
reshape long week, i(id) j(visit)
sort id visit
list in 1/5
* Plot in fig3.1 in DHLZ
** wrong spaghetti plot **
twoway line week visit, s(i) ytitle(weight (kg)) xtitle(time (weeks))
** correct spaghetti plot **
twoway line week visit, connect(ascending) s(i) ytitle(weight (kg)) xtitle(time (weeks))

*To obtain Fig3.2 in DHLZ:

reshape wide
*One way of generating the standardized weight  
summarize week1
gen sweek1=(week1-25.02)/2.47

* A simpler way:
drop sweek1
egen sweek1=std(week1)
egen sweek2=std(week2)
egen sweek3=std(week3)
egen sweek4=std(week4)
egen sweek5=std(week5)
egen sweek6=std(week6)
egen sweek7=std(week7)
egen sweek8=std(week8)
egen sweek9=std(week9)

reshape long week sweek, i(id) j(visit)
sort id visit
twoway line sweek visit, connect(ascending) s(i) ytitle(Residual) xtitle(Time (weeks))

*Linear regression;
regress week visit

*Close the log file
log close

exit, clear