Forum: open-discussion

RE: Cross generational analysis [ Reply ]
By: Will Beasley on 2013-11-22 05:08

NlsyAce.pdf (264) downloads

Hi John, see if the new vignette example (number 6) does what you want. The only difference is that you'll need to stack dsOutcomes yourself for Step 5.

Read the Gen1 and Gen2 CSVs into R individually, and you'll have something like dsOutcomesGen1 and dsOutcomesGen2. Before you stack them on top of each other, I recommend trimming out all the unused variables, to keep it more manageable.

The function `rbind.fill()` in plyr makes it easy to stack the datasets, as long as the variable names are the same.

As I'm writing this, I see that this probably would be helpful to be a new vignette example, but my embedded example CSVs don't have a good variable for this. Realistically, it will be a few weeks before I'll get back into this and add it. Tell us if you have questions in the meantime.

RE: Cross generational analysis [ Reply ]
By: Will Beasley on 2013-11-15 07:05

[forum:40041]

John, we haven't forgotten about you. I've been adding a lot of stuff to the vignettes. The version on CRAN hasn't changed, but the R-Forge versions have changed.

A complete example is one of the upcoming additions for this weekend. It won't be complicated, because the linking datasets and `SubjectDetails79` already contains both generations. It's just an issue of joining the two extracts, and avoiding an inner join (which would drop people).

Cross generational analysis [ Reply ] By: John Fuerst on 2013-10-31 20:05	[forum:39976]
Hello, I was wondering if it would be possible to conduct a cross generational analysis using the NlsyLinks package. If so, would it be possible for someone to explain: (a) how to create a workable merged (Gen1 + Gen2) .csv file and (b) how to get this .csv file to read into NlsyLinks?