Author: Robin Denz
simDAG is an R-Package which can be used to generatedata from a known directed acyclic graph (DAG) with associatedinformation on distributions and causal coefficients. The root nodes aresampled first and each subsequent child node is generated according to aregression model (linear, logistic, multinomial, cox, …) or otherfunction. The result is a dataset that has the same causal structure asthe specified DAG and by expectation the same distributions andcoefficients as initially specified. It also implements a comprehensiveframework for conducting discrete-time simulations in a similarfashion.
A stable version of this package can be installed from CRAN:
install.packages("simDAG")and the developmental version may be installed from github using theremotes R-Package:
library(remotes)remotes::install_github("RobinDenz1/simDAG")If you encounter any bugs or have any specific feature requests,please file anIssue.
Suppose we want to generate data with the following causalstructure:
whereage is normally distributed with a mean of 50 anda standard deviation of 4 andsex is bernoulli distributedwithp = 0.5 (equal number of men and women). Both of these“root nodes” (meaning they have no parents - no arrows pointing intothem) have a direct causal effect on thebmi. The causalcoefficients are 1.1 and 0.4 respectively, with an intercept of 12 and asigma standard deviation of 2.death is modeled as abernoulli variable, which is caused by bothage andbmi with causal coefficients of 0.1 and 0.3 respectively.As intercept we use -15.
The following code can be used to generate 10000 samples from thesespecifications:
library(simDAG)dag<-empty_dag()+node("age",type="rnorm",mean=50,sd=4)+node("sex",type="rbernoulli",p=0.5)+node("bmi",type="gaussian",formula=~12+ age*1.1+ sex*0.4,error=2)+node("death",type="binomial",formula=~-15+ age*0.1+ bmi*0.3)set.seed(42)sim_dat<-sim_from_dag(dag,n_sim=100000)By fitting appropriate regression models, we can check if the datareally does approximately conform to our specifications. First, letslook at thebmi:
mod_bmi<-glm(bmi~ age+ sex,data=sim_dat,family="gaussian")summary(mod_bmi)#>#> Call:#> glm(formula = bmi ~ age + sex, family = "gaussian", data = sim_dat)#>#> Coefficients:#> Estimate Std. Error t value Pr(>|t|)#> (Intercept) 11.89194 0.07954 149.51 <2e-16 ***#> age 1.10220 0.00158 697.41 <2e-16 ***#> sexTRUE 0.40447 0.01268 31.89 <2e-16 ***#> ---#> Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1#>#> (Dispersion parameter for gaussian family taken to be 4.022026)#>#> Null deviance: 2361465 on 99999 degrees of freedom#> Residual deviance: 402190 on 99997 degrees of freedom#> AIC: 422971#>#> Number of Fisher Scoring iterations: 2This seems about right. Now we look atdeath:
mod_death<-glm(death~ age+ bmi,data=sim_dat,family="binomial")summary(mod_death)#>#> Call:#> glm(formula = death ~ age + bmi, family = "binomial", data = sim_dat)#>#> Coefficients:#> Estimate Std. Error z value Pr(>|z|)#> (Intercept) -14.6833 3.5538 -4.132 3.6e-05 ***#> age 0.2607 0.1698 1.535 0.125#> bmi 0.1842 0.1402 1.314 0.189#> ---#> Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1#>#> (Dispersion parameter for binomial family taken to be 1)#>#> Null deviance: 258.65 on 99999 degrees of freedom#> Residual deviance: 214.03 on 99997 degrees of freedom#> AIC: 220.03#>#> Number of Fisher Scoring iterations: 13The estimated coefficients are also very close to the ones wespecified. More examples can be found in the documentation and themultiple vignettes.
If you use this package, please cite the associated article:
Denz, Robin and Nina Timmesfeld (2025). Simulating ComplexCrossectional and Longitudinal Data using the simDAG R Package. arXivpreprint, doi: 10.48550/arXiv.2506.01498.
© 2024 Robin Denz
The contents of this repository are distributed under the GNU GeneralPublic License. You can find the full text of this License in thisgithub repository. Alternatively, seehttp://www.gnu.org/licenses/.