Categorizing a date variable

[Harish Kumar]

Hi,

I have a date variable dov (date of visit) in float %td format. The time period ranges from 2015 to 2018. I want to generate a time variable which will look like year1, year2, year3 only. Kindly help me in this.

dov

19-Jul-15

1-May-15

1-May-15

10-Jun-15

1-Aug-15

20-Aug-15

15-Mar-16

2-Dec-16

3-Dec-18

23-Dec-16

10-Feb-17

2-Dec-16

3-Dec-16

2-Dec-16

With Thanks
Harish

[Amit Narnoli]

Try the following code

gen numdate = substr(dov, strrpos(var1, "-")+1,.)
destring numdate, replace
egen mindate = min(numdate)
gen numyear= 1 if numdate==mindate
sort numyear
replace numyear = numdate[_n]-numdate[1]+1 if _n>1
tostring numyear, replace
gen year = "year"+numyear
destring numyear, replace
drop numdate mindate

You will get two variables
1. numyear - representing year in numeric form
2. year - representing year in year1, year2..form as you want

[Harish Kumar]

Thanks for this.
When I run the first line of the codes, it is showing as var1 not found. Could you please check if var1 required?

[Joro Kolev]

I do not understand the task. How would this new year variable look for example for dates 10-Jun-15 and 23-Dec-16?

Otherwise you can extract easily components from td variable, e.g.,

gen year = year(dov)

will extract only the year in which the visit took place.

[Felix Scholl]

You want a categorical variable that reflects the year of the visit (2015, 2016, 2017? Please note that you should provide a data example using dataex (see in the FAQ). It is difficult to help you without knowing how your data looks.

[Harish Kumar]

I do not understand the task. How would this new year variable look for example for dates 10-Jun-15 and 23-Dec-16?

Otherwise you can extract easily components from td variable, e.g.,

gen year = year(dov)

will extract only the year in which the visit took place.

Thanks a lot. It worked.

[Amit Narnoli]

Thanks for this.
When I run the first line of the codes, it is showing as var1 not found. Could you please check if var1 required?

Sorry for this. Replace var1 by dov.
[CODE]

[Harish Kumar]

You want a categorical variable that reflects the year of the visit (2015, 2016, 2017? Please note that you should provide a data example using dataex (see in the FAQ). It is difficult to help you without knowing how your data looks.

Thanks for this, I will follow it.

[Harish Kumar]

Hi,

I have three variables in the attached data (date, PatientID, followup). I want to calculate the number of days between followup visits for each patient. Some patients had no followups, and some had 1,2,3,7,12 ... followups. Your help will be much appreciated.

[Nick Cox]

Please read and act on https://www.statalist.org/forums/help#stata which explains why ,dta attachments are not a good idea.

Also, the question has nothing to do with the thread title. When that's true you should start a new thread.