Stata code wighted census data for PCA

[Liton Chakraborty]

Hello all,

I am new to the Stata Community and working on the 2016 Canadian census of population micro dataset to create a social deprivation index using census tract level data on socio-economic characteristics of Canadians. While preparing the required variables for PCA in Stata, I found that the dataset came up with a person weight variable (COMPW2) and STATCAN has recommended using COMPW2 to represent the population in any further statistical analysis. COMPW2 weight is the same for all household members but varies across households, ranges from 1 to 17, with an average value of approximately 4. I have a focus on a particular CMA and the analysis will be done at the census tract level.

Would you please comment whether I have correct Stata commands while generating new variables using the COMPW2 weight variable.

Variable: percentage/proportions of females in each census tract

Stata Code:
*total number of females by census tract
egen nofem=total COMPW2 if sex==1, by(censustract)
alternatively, bysort censustract: egen nofem=total COMPW2 if sex==1
*total population by census tract
egen nopop=total COMPW2, by(censustract)
*proportion of females by census tract
egen propfem=nofem/nopop

Your help will be appreciated. Thanks, Liton

[William Lisowski]

Welcome to Statalist, Liton.

You have accidentally posted your topic in Statalist's Mata Forum, which is used for discussions of Stata's Mata language. Your question will see a much larger audience if you post it in Statalist's General Forum.

Also, if you have not already done so, take a look at the Statalist FAQ linked to at the top of this page for posting guidelines and suggestions.

[Liton Chakraborty]

My apologies, William. I will repost it in the general forum. Many thanks.