Simulated data set

Share:

Description

For testing purposes we constructed the very extreme and unbalanced simulated binomial data set simul. The pattern of this data set is typical of models for rare events, e.g. rare diseases or financial defaults. Based on a fixed number of dims = 10 covariates consisting of nine binary variables and the intercept, the design matrix X is built by computing all 2^9 possible 0/1 combinations. The true parameter vector is beta={0.05,2,1.5,-3,-0.01,-1.3,2.9,-2.1,0.5,-0.2}. For details concerning the simulation of the data set see the paper by Fussl, Fruehwirth-Schnatter and Fruehwirth (2013). To use the data set with the function IndivdRUMIndMH, binary outcomes are reconstructed from the binomial observations and saved as simul_binary.

Usage

1
2
3

Format

The binomial data set simul consists of 512 binomial observations and the following 12 variables:

yi

number of successes for each covariate pattern

Ni

group size for each covariate pattern

X,X.1,...,X.8

binary covariates

X.9

intercept

Only 490 covariate patterns have a group size Ni > 0 and will be included when using the functions dRUMIndMH, dRUMAuxMix and dRUMHAM.

The binary data set simul_binary consists of 25803 binary observations and the following 11 variables:

y

binary response variable

X,X.1,...,X.8

binary covariates

X.9

intercept

Source

Agnes Fussl, Sylvia Fruehwirth-Schnatter and Rudolf Fruehwirth (2013), "Efficient MCMC for Binomial Logit Models". ACM Transactions on Modeling and Computer Simulation 23, 1, Article 3, 21 pages.

See Also

dRUMIndMH, IndivdRUMIndMH

Examples

1
2
3
4
  data(simul)
  data(simul_binary)
  ## see dRUMIndMH and IndivdRUMIndMH documentation for examples using
  ## these data