csmpv (version 1.0.5)

rms_model: A Wrapper for Building Predictive Models using the rms Package

Description

This wrapper function harnesses the capabilities of the rms package to construct robust predictive models using various modeling techniques. Whether you're working with linear regression (lm), generalized linear models (glm), or Cox proportional hazards models (coxph) directly or generating these objects from other functions, this function streamlines the model-building process and enhances analysis. The function features built-in bootstrap-based internal validation, model score generation, and result storage.

Usage

rms_model(afit, data = NULL, newdata = NULL, newY = FALSE, u = 2, outfile)

Value

Predictions are returned.

Arguments

afit

A model fit from lm, glm, or coxph.

data

A data frame used to obtain the 'afit' object. This parameter is only required when the outcome variable is time-to-event; otherwise, the data is extracted from 'afit'.

newdata

A new data frame for prediction.

newY

A logical variable indicating whether 'newdata' contains the outcome variable.

u

A single numeric value representing follow-up time for survival outcomes, used to estimate the survival probability for the given time point.

outfile

A string indicating the output file name without the file type extension, but including the complete path information.

Author

Aixiang Jiang

Details

The wrapper function capitalizes on the power of the rms package, providing a seamless interface for creating predictive models. With the rms package, the function conducts efficient bootstrap-based internal validation, enabling thorough model evaluation. Nomograph plots and C-index calculations are also at your disposal. Furthermore, the function adapts to your needs by predicting model scores for new datasets. In the absence of new data, it generates predictions based on the training data. However, when you provide an independent dataset complete with outcome variable information, the function extends to external validation. This enables assessing model performance in an independent context using the rms R package.

All results, from the model itself to internal validation metrics and predictions, are conveniently stored locally for easy access and further analysis.

References

Harrell Jr F (2023). rms: Regression Modeling Strategies. R package version 6.7-1, <https://CRAN.R-project.org/package=rms>

Harrell Jr F (2023). Hmisc: Harrell Miscellaneous. R package version 5.1-1, <https://CRAN.R-project.org/package=Hmisc>

Examples

# Load in data sets:
data("datlist", package = "csmpv")
tdat = datlist$training
vdat = datlist$validation

# The function saves files locally. You can define your own temporary directory. 
# If not, tempdir() can be used to get the system's temporary directory.
temp_dir = tempdir()
# As an example, let's define Xvars, which will be used later:
Xvars = c("highIPI", "B.Symptoms", "MYC.IHC", "BCL2.IHC", "CD10.IHC", "BCL6.IHC")

# The function can work with multiple models and multiple outcome types. 
# Here, we use continuous as an example:
clr = LASSO2_reg(data = tdat, biomks = Xvars,
                 outcomeType = "continuous", Y = "Age",
                 outfile = paste0(temp_dir, "/continuousLASSO2_reg"))
pclr = rms_model(clr$fit, newdata = vdat,
                 outfile = paste0(temp_dir, "/pred_LASSO2reg_continuous"))
# You might save the files to the directory you want.

# To delete the "temp_dir", use the following:
unlink(temp_dir)