*cd "C:\Desktop\PBC_Analysis"
cd "/home/scottt/Desktop/"
log using pbclog, clear

use pbc.dta, clear

describe
codebook

* Create new continuous variables
generate ageyrs = age/365.25
generate fuyrs = fudays/365.25
describe

* Create new categorical variable
generate censored = .
replace censored = 1 if status == 0 | status == 1
replace censored = 2 if status == 2

* Add labels to new variables
label variable ageyrs "Age (yrs)"
label variable fuyrs "Follow-up (yrs)"
label variable censored "2-level survival status"

* Overwrite some existing labels
label variable age "Age (days)"
label variable fudays "Follow-up (days)"
label variable status "3-level survival status"

* Add value labels
label define statuslabel 0 "Censored" 1 "Censored due to liver treatment" 2 "Dead"
label value status statuslabel

label define censoredlabel 1 "Censored" 2 "Dead"
label value censored censoredlabel

label define druglabel 1 "D-penicillamine" 2 "Placebo"
label value drug druglabel

label define sexlabel 1 "Female" 2 "Male"
label value sex sexlabel

label define asciteslabel 0 "No" 1 "Yes"
label value ascites asciteslabel

label define stagelabel 1 "1" 2 "2" 3 "3" 4 "4"
label value stage stagelabel

codebook
list in 1/10

* just continuous variables, just # non-missing obs, mean, SD, min, and max
summarize fudays fuyrs age ageyrs bili chol album
* detail = percentiles (1, 5, 10, 25, 40, 75, 90, 95, 99), 4 smallest, 4 largest, n obs, mean, SD, variance, skewness, kurtosis
summarize fudays fuyrs age ageyrs bili chol album, detail
summarize ageyrs, meanonly
summarize ageyrs if sex == 1
summarize ageyrs if sex == 2

tabulate sex
tabulate drug, missing
tabulate sex drug, missing column
* sex in rows, drug in columns

tabstat ageyrs, by(sex) statistics(count mean sd p25 median p75) missing
tabstat ageyrs, by(drug) statistics(count mean sd p25 median p75) missing


table stage sex drug, missing
* stage in rows, sex in (nested) columns, drug in "super"-columns

table stage sex drug, contents(freq mean ageyrs) missing