MRPC (version 2.0.0)

SimulationDemo: Same data but different node orderings.

Description

Investigate the impact of the same data but with different node orderings across three methods: MRPC (Badsha et al., 2018), the mmhc method (implemented in bnlearn; Scutari, 2010) and the pc method (implemented in pcalg; Kalisch et al., 2012).

Usage

SimulationDemo(N, model, signal, n_data, n_nodeordering)

Arguments

N

The number of observation.

model

Either 'truth1' or 'truth2' to specify the model you want to generate data from.

signal

The signal strength which is the coefficient of the parent nodes in the linear model.

n_data

The number of data sets you want to generate for testing the results of the three methods.

n_nodeordering

The number of times you want to reorder the nodes for testing the results of the three methods.

Value

Matrix

Details

The output is a matrix where each row indicates the data sets and the columns are the results of the different node orderings by three methods MRPC, mmhc, and pc respectively.

References

1.Md. Bahadur Badsha, Audrey Qiuyan Fu: Learning causal biological networks with generalized Mendelian randomization.bioRxiv (2017), 171348.

2.Marco Scutari (2010). Learning Bayesian Networks with the bnlearn R Package. Journal of Statistical Software, 35(3), 1-22.

3.Kalisch, M., Machler, M., Colombo, D., Maathuis, M.H. & Buhlmann, P. Causal Inference Using Graphical Models with the R Package pcalg. J. Stat.Softw. 47, 26 (2012).

Examples

# NOT RUN {
# We want to generate 2 different data sets for truth1 with
# signal = 1, N = 100 and 6 different node orderings. Therefore,
# you will get 2 by 18 output matrix, where first and second row 
# are the data 1 and 2 respectively. Column 1:6 is the result of 
# MRPC with node order 1, 2, ..., 6, then column 7:12 is for mmhc and 
# column 13:18 is for pc.

library(bnlearn) # for mmhc
library(pcalg)   # for pc

# Run
Output <- SimulationDemo(N = 100, 'truth1',
                         signal = 1, n_data = 2,
                         n_nodeordering = 6)

# }