Hi all,
I am having some trouble with replicability using the corr2data command to simulate autocorrelated data.
Basically, I want to perform 100 replications of simulated datasets with autocorrelated variables, each replication using a seed going from 1 to 100. However, I have noticed that corr2datais obtaining exactly the same data from consecutive even-odd seeds (i.e. the data from seed=2 is the same as the data from seed=3, and this happens with seeds 4-5, 6-8, ...etc). I attach a photo with the pvalue of a variable in a regression to illustrate this case, along with the number of repetition and the corresponding seed.
Has anyone noticed this and/or why does this occur? The obvious solution for me is to use other seed structure, but I still don't find this normal and want to figure out if its a programming error.
Thanks,
Christian
Array
Related Posts with Inconsistency with corr2data command
Cox proportional hazards model a non constant covariateDear stata listers I'm using stata to analyse time-to-event data. Specifically, I'm trying to relat…
Timeseries - cumulative impactI have been using Regression with Newey-West standard errors Code: newey for time-series regressi…
IVQREG2: module to estimare structural quantile functions now available in SSCDear All: Thanks to Kit Baum, ivqreg2 is now available in SSC. This module estimates the structura…
Help to import from ExcelHi. I am trying to import an excel file into Stata. However, when I use the option firstrow, Stata …
How to create a dummy variable from two variablesHi! I have a dataset with two variables for test scores: scores in 2002 and scores in 2013. I want n…
Subscribe to:
Post Comments (Atom)
0 Response to Inconsistency with corr2data command
Post a Comment