3

I'm building an ERGM on statnet and one of the lines shows NA, which I don't know how to interpret. All of the other nodefactors show results but this one just shows:

nodefactor.religion.1                 NA     0.0000      0      NA       NA

The column it's pulling from is just a bunch of 1s and 0s, so I don't see where the problem is.

Filburt
  • 17,626
  • 12
  • 64
  • 115
zacharay
  • 31
  • 1
  • hard to tell without more details. could be due to collinearity. could you provide more details about your network, the full ERGM output, and a [minimal reproducible example](https://stackoverflow.com/help/mcve)? – paqmo Apr 10 '20 at 13:46
  • @paqmo thanks! The network has 206 vertices and about 600 edges. The vertex attributes are pulled from a meta, and the ERGM gives results for all of the attributes except "religion". Estimate Std. Error MCMC % z value Pr(>|z|) edges -5.8087 0.3743 0 -15.518 <1e-04 *** nodefactor.caste.SCHEDULE TRIBE 0.5634 0.1114 0 5.059 <1e-04 *** nodefactor.religion.1 NA 0.0000 0 NA NA – zacharay Apr 10 '20 at 15:05
  • yeah, hard to tell without seeing the data. Can you share it, or a subset of it, using `dput()` in your question? You could try examining the mixing matrix for religion to see if you spot something unusual--`mixingmatrix(network, "religion")` – paqmo Apr 10 '20 at 15:23

0 Answers0