Skip to content

Latest commit

 

History

History
177 lines (142 loc) · 5.08 KB

README.md

File metadata and controls

177 lines (142 loc) · 5.08 KB
Raw

{flexFitR}

{flexFitR} is an R package designed for efficient modeling and analysis of large and complex datasets. It offers powerful tools for parameter estimation, model fitting, and visualization, leveraging the {optimx} package for optimization and the {future} package for parallel processing.

Installation

You can install the development version of flexFitR from GitHub with:

# install.packages("devtools")
devtools::install_github("AparicioJohan/flexFitR")

Features

  • Parameter Estimation: Utilizes {optimx} and derivative-free algorithms to solve and estimate parameters for a given function.
  • Parallelization: Implements parallel processing using the {future} package, enabling efficient fitting of hundreds of curves simultaneously.
  • Visualization Tools: Provides a variety of plots to visualize model fits, correlations, predictions, and more.
  • Statistical Rigor: Offers standard errors and p-values for coefficients, as well as for predictions, supporting robust conclusions and interpretations.
  • Prediction: Supports diverse prediction types, including point predictions, area under the curve (AUC), first and second derivatives, user-defined parameter functions, and more advanced transformations or custom expressions based on model parameters.
  • Flexibility: Allows users to fix certain parameters in the model and specify different initial values per grouping factor.
  • Custom Modeling Functions: Equipped with built-in modeling functions for common analysis tasks, while also permitting users to supply their own custom functions.

Example

Here’s a simple example to get you started with {flexFitR}. This example demonstrates fitting a piecewise regression model:

library(flexFitR)

dt <- data.frame(
  time = c(0, 29, 36, 42, 56, 76, 92, 100, 108),
  variable = c(0, 0, 0.67, 15.11, 77.38, 99.81, 99.81, 99.81, 99.81)
)
plot(explorer(dt, time, variable), type = "xy")

plot xy

fun <- function(t, t1 = 45, t2 = 80, k = 0.9) {
  if (t < t1) {
    y <- 0
  } else if (t >= t1 && t <= t2) {
    y <- k / (t2 - t1) * (t - t1)
  } else {
    y <- k
  }
  return(y)
}
# Fitting piecewise regression
mod_1 <- dt |>
  modeler(
    x = time,
    y = variable,
    fn = "fun",
    parameters = c(t1 = 45, t2 = 80, k = 90)
  )
print(mod_1)

Call:
variable ~ fun(time, t1, t2, k) 

Residuals:
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
0.00000 0.00000 0.00000 0.07444 0.00000 0.67000 

Optimization Results `head()`:
 uid   t1 t2    k   sse
   1 38.6 61 99.8 0.449

Metrics:
 Groups      Timing Convergence Iterations
      1 0.4909 secs        100%   511 (id)
# Auto plot
plot(mod_1)

plot fin

# Coefficients
coef(mod_1)
# A tibble: 3 × 6
    uid coefficient solution std.error `t value` `Pr(>|t|)`
  <dbl> <chr>          <dbl>     <dbl>     <dbl>      <dbl>
1     1 t1              38.6    0.0779      496.   4.54e-15
2     1 t2              61.0    0.0918      665.   7.82e-16
3     1 k               99.8    0.137       730.   4.47e-16
# Variance-Covariance Matrix
vcov(mod_1)
$`1`
              t1           t2            k
t1  6.061705e-03 -0.002940001 1.877072e-07
t2 -2.940001e-03  0.008431400 4.204939e-03
k   1.877072e-07  0.004204939 1.870426e-02
# Making predictions
predict(mod_1, x = 45)
# A tibble: 1 × 4
    uid x_new predicted.value std.error
  <dbl> <dbl>           <dbl>     <dbl>
1     1    45            28.5     0.223

Documentation

For detailed documentation and examples, visit flexFitR

Contributing

Contributions to flexFitR are welcome! If you’d like to contribute, please fork the repository and submit a pull request. For significant changes, please open an issue first to discuss your ideas.

Code of Conduct

Please note that the flexFitR project is released with a Contributor Code of Conduct. By contributing to this project, you agree to abide by its terms.

License

flexFitR is licensed under the MIT License. See the LICENSE file for more details.