The DHS Program User Forum

Today's Messages (on) | Unanswered Messages (off)

Forum: India

Wed, 29 June 2022 06:14

hvs0013
Messages: 2
Registered: April 2022

Member

Hi,

I am trying to merge the PR and the BR datasets for India. I used the variables v001, v002, b16 in the BR file and hv001, hv002, hvidx in the PR file to merge the data. Furthermore, I dropped b16=0 and b16=. before the merge. However, when I try to perform the merge, some observations from the BR file do not get merged to the PR file. I understand that many observations in the PR file will not get merged to the PR file as some people born have died or are not living in the household. But I am confused regarding why some observations in the BR file will not get merged to the PR file. Thank you

Report message to a moderator

Forum: General

Topic: Interpolated surfaces

Re: Interpolated surfaces [message #24729 is a reply to message #24651]

Wed, 29 June 2022 09:00

Rose-DHS
Messages: 2
Registered: October 2021

Member

To get the displaced GPS coordinates of the clusters, you'll need to download the GE dataset(s) for the survey(s) of interest.

You can link the cluster number (V001) variable from the recode file with the DHSCLUST variable from the geographic dataset.

Could you provide us with more information to help us better answer your question?

Report message to a moderator

Forum: Bangladesh

Topic: Caseid variable is not available in the height-weight data for BDHS 2004

Re: Caseid variable is not available in the height-weight data for BDHS 2004 [message #24728 is a reply to message #3043]

Wed, 29 June 2022 08:30

Wolf
Messages: 1
Registered: June 2022

Member

Thanks for your help above.

Please could you update this in the merge instructions? To reflect that it is not the 'midx' variable in the child recode that is equivalent to 'HWLINE' in the new file, it is instead 'b16'.

Just to confirm, in this case I will need to drop the last digit of the 'caseID' in the child recode, and instead replace it with 'b16' so that the codes will match in both files and can then be merged together?

Report message to a moderator

Re: Caseid variable is not available in the height-weight data for BDHS 2004 [message #24731 is a reply to message #24728]

Wed, 29 June 2022 09:50

Trevor-DHS
Messages: 772
Registered: January 2013

Senior Member

You are really creating the household ID which is the first 12 characters of the caseid (effectively dropping the woman's line number - the last 3 characters of caseid [not just the last digit]). You are the matching on the household ID and the household line number in the two datasets so that the two datasets can be merged.

Report message to a moderator

Forum: Fertility

Topic: R code for TFR estimation -SAE perspective

R code for TFR estimation -SAE perspective [message #24732]

Wed, 29 June 2022 10:18

DIO
Messages: 2
Registered: March 2020

Member

I am trying to calculate TFR in DHS.rates using Kenya DHS data 2014 with purpose of attempting further modelling in small area estimation (SAE). However I am getting the following error:

> setwd("~/scountytfr")
> library(DHS.rates)
> library(haven)
> KEIR2014sbset <- read_dta("C:/Users/DAVID/Documents/scountytfr/KEIR2014sbset.dta ")
> fert(KEIR2014sbset,Indicator = "tfr", JK = "Yes", Class="Sub_Countyn")

The current function calculated TFR based on a reference period of 36 months
The reference period ended at the time of the interview, in 2014.58 OR May - Oct 2014
The average reference period is 2013.08
Error in matrix(0, nrow = max(as.numeric(Dat$DomID)), ncol = 10) :
invalid 'nrow' value (too large or NA)

What to do?

Report message to a moderator

Current Time: Wed Jun 29 21:28:37 Coordinated Universal Time 2022