magic_irri

Multiparent advanced generation inter-cross (MAGIC) populations are
ideal for learning complex models because of their high genetic
recombination, diversity and large sample size

Model developed as an example of multiple trait modelling in plant genetics for the invited talk "Bayesian Networks, MAGIC Populations and Multiple Trait Prediction" delivered by Marco Scutari at the 5th International Conference on Quantitative Genetics (ICQG 2016).

From the library bnlearn, we used the next code in R (to reproduce https://openreview.net/pdf?id=fPVRcJqspu):

install.packages("bnlearn");
library(bnlearn)
data(package="bnlearn")

# Load the RDA file (this creates an object named 'bn' in your environment)
load("magic-irri.rda") # Replace with your actual file path

# Rename the object for clarity
magic_bn <- bn

# Load the RDS file directly
magic_bn <- readRDS("magic-irri.rds") # Replace with your actual file path

# Generate 1000 synthetic observations from the network
magic_data <- rbn(magic_bn, n = 2000)

# Save as CSV
write.csv(magic_data, "magic_irri_synthetic_data.csv", row.names = FALSE)

For this dataset we decided to define the target variable as "HT". This is an arbitrary choice, and you may want to explore other columns as the target variable. The target variable is the variable that you want to predict in your machine learning model.