implement umxDiscTwin discordant twin test of causation

[tbates]

Function call:

tmp = umxDiscTwin(x = "exercise", y = "depression", data= twinData, mzZygs = c("MZFF", "MZMM"), dzZygs = c("DZFF", "DZMM", "DZOS"))

Internal modelling

MZ1  = lme(fixed = y ~ x + FamMeanX, random = ~ 1|FAMID, data = umx_scale(MZ), na.action = "na.omit")

• [x] take raw data, parse by zygosity, make long, understand FAMID
• [x] getting the graphics working
• [x] Make a graph of all three groups (pop, discordant DZ and discordant MZ).
• [x] Building in lme code
• [x] Compute B_within, B_between (and CIs) vialme() with random intercept for FAMID + control intrapair-mean.
• [x] Improve graphics
• [x] In the help, show user the model used, with fixed and random effects formulae.
• [ ] Show user all "working results": i.e., print out the 'lme' models (use umxAPA)

Refs

• McGue, M., Osler, M., & Christensen, K. (2010). Causal Inference and Observational Research: The Utility of Twins. Perspectives on Psychological Science, 5, 546-556. doi:10.1177/1745691610383511
• Begg, M. D., & Parides, M. K. (2003). Separation of individual-level and cluster-level covariate effects in regression analysis of correlated data. Stat Med, 22(16), 2591-2602. doi:10.1002/sim.1524

[tbates]

Implement the discordant twin design for testing causality. Create bar graph of r in the population, MZ, and DZ groups, with CI, and return correlations as a table also.

McGue, M., Osler, M., & Christensen, K. (2010). Causal Inference and Observational Research: The Utility of Twins. Perspectives on Psychological Science, 5, 546-556. doi:10.1177/1745691610383511