***************** INTRODUCTION TO STATA ************

***** 11 November 2022 *************

***** EXERCISE
*IMPORT data "01 - Body Data"STATA code: import delimited "C:\...\01 - Body Data.txt"Data Set : Body Data - Body and exam measurements from 300 subjects 
*Describe pulse variable, are there any extreme values?
*Find the mean, standard deviation and quartiles values
*Plot distribution of pulse rate by histogram and boxplot
*Describe sex
*Plot distribution of pulse rate by sex

// IMPORT
import delimited "C:\...\01 - Body Data.txt"
//Data Set 1: Body Data
//Body and exam measurements are from 300 subjects 

*from drop-down menu: Import text data (delimited , .csv,...)

summarize pulse,d
histogram pulse
graph box pulse

tabulate gender1m
 graph box pulse, over(gender1m)
 histogram pulse, by(gender1m)

*identify extreme values by Tuckey definition
display 80-64
display 64 - 16*1.5
display 80 + 16*1.5

 list if pulse<40
 
 list if pulse>104
 
 
***** DO files

*	You can type all the same commands into the Do-file that you would type into the command windo
*	BUT…the Do-file allows you to save your commands
*	Your Do-file should contain ALL commands you executed – at least all the "correct" commands!
*	I recommend never using the command window or menus to make CHANGES to data
*	Saving commands in Do-file allows you to keep a written record of everything you have done to your data 
   *	Allows easy replication
   *	Allows you to go back and re-run commands, analyses and make modifications

   
   ******** GAUSSIAN APPROXIMATION ********************
summarize pulse
histogram pulse, normal bin(10)

* which is the range that include nearly the central 68% of the observation?
	display 71.76667-12.12803
	display 71.76667+12.12803
	 count if pulse>59.63864 & pulse<83.8947
	 display 205/300
 
 * which is the range that include nearly the central 95% of the observation?
	 display 71.76667+2*12.12803
	*96.02273
	display 71.76667-2*12.12803
	*47.51061
	count if pulse>47.51061 & pulse<96.02273
	*  288
	display 288/300
	*.96 bin(10)

* find the 99th percentile
summarize pulse,d 
*100
display invnormal(0.99)
display 2.3263479*12.12803+71.76667
*99.980687


   
******************** WORK WITH BODY DATA (Triola et al Appenix B -data set 1)***********************
// IMPORT
import delimited "C:\...\01 - Body Data.txt"
//Data Set 1: Body Data
//Body and exam measurements are from 300 subjects 

*from drop-down menu: Import text data (delimited , .csv,...)

//variables, data editor
edit

* decription of data 
codebook
summ
inspect
 
// add an identificator variable
gen id = _n,before(age)
label variable id "unique identific" 

 save "C:\...bodydata.dta"

******* QUALITATIVE VARIABLES **********
tabulate gender1m
 label define Sex 1 "Male" 0 "Female"
 label values gender1m Sex
 decode gender1m,generate(sex)

graph pie, over(gender1m) title(Sex distribution) legend(on order(1 "Males" 2 "Females" ))
graph pie, over(sex) title(Sex distribution) 

  graph bar, over(gender1m) 
  graph bar, over(sex) 
  *customise titles with drop down menu 
  
 graph bar, over(gender1m) 

 graph bar (count), over(gender1m) 
  
  
******* QUANTITATIVE VARIABLES **********
 tabulate bmi
* not readable!
 ********** histogram
histogram bmi 
	*customise number of classes by drop-down menu
 
*create classes of width 1:
 egen float bmi_c = cut(bmi), at(15(1)60) 
 tabulate bmi_c
	*see data change 
 
 ********** Galton ogiva
  sort bmi_c 
  cumul bmi_c, gen(cu_bmi_c)
  line cu_bmi_c bmi_c
	
 ********** Boxplot
  graph box bmi	
  
  
  ********* summary indicatoris
  summarize bmi
 
 *inspect extreme data
  list if bmi>50
  sum waist
  sum armcirc
  
  
  ****** plots by sex 
  
histogram bmi, normal by(gender1m)
by gender1m, sort : summarize bmi, detail
graph box bmi,by(sex)	
graph box bmi,over(sex)	


 **** classify by BMI categories
 
egen bmi_class= cut(bmi), at(0,18.5,25,30,40,100)
 graph box bmi,over(bmi_class)	
 //generate bmi_class2 = bmi
//recode   bmi_class2 (min/18.5=1) (18.5/25=2) (25/max=3)
 
 graph box waist,over(bmi_class)