- #BENCHMARK TESTING EXAMPLES FULL#
- #BENCHMARK TESTING EXAMPLES CODE#
- #BENCHMARK TESTING EXAMPLES SERIES#
For each combination of parameters, we run the microbenchmark function from the package microbenchmark 1000 times for each implementation and only save the median benchmark time of these 1000. For the preliminary benchmark testing, we will input the response times as a vector to each function instead of inputting individual response times for two reasons: first, this is the most common way to input the response time data and second, this allows for the R-based implementation to exploit vectorization to make that implementation as fast as possible.
#BENCHMARK TESTING EXAMPLES CODE#
To achieve this rigor, we define a parameter space (in a code chunk below) and loop through each combination of parameters. We want to determine the performance of each implementation across a variety of parameters in order to identify any slow areas for individual implementations. Please note that we will not be testing the functions for accuracy or consistency in this vignette, as that is covered in the Validity Vignette.īenchmarking the Density Function Approximations The second section will record the benchmark data of parameter estimation that uses the density function approximations in the optimization process for fitting to real-world data this section will also include visualizations to illustrate the differences among the density function approximations.
#BENCHMARK TESTING EXAMPLES SERIES#
The first section will record the benchmark data of our implementations of the density function approximations and show the results through a series of visualizations. To this aim, we present benchmark data of currently available approximation methods from the literature, our streamlined implementations of those approximation methods, and our own novel approximation method.
This vignette is designed to compare the speed of the currently available implementations that calculate the lower probability density of the DDM. However, the density function for the DDM is notorious for containing an unavoidable infinite sum hence, the literature has produced a few different methods of approximating the density. Since the DDM is widely used in parameter estimation usually involving numerical optimization, significant effort has been put into making the evaluation of its density as fast as possible.
Our implementation of the DDM has the following parameters: \(a \in (0, \infty)\) (threshold separation), \(v \in (-\infty, \infty)\) (drift rate), \(t_0 \in [0, \infty)\) (non-decision time/response time constant), \(w \in (0, 1)\) (relative starting point), \(sv \in (0, \infty)\) (inter-trial-variability of drift), and \(\sigma \in (0, \infty)\) (diffusion coefficient of the underlying Wiener Process). Two examples of using dfddm for parameter estimation are provided in the Example Vignette. An empirical validation of the implemented methods is provided in the Validity Vignette. An overview of the mathematical details of the different approximations is provided in the Math Vignette.
#BENCHMARK TESTING EXAMPLES FULL#
Generating Benchmark Data for Parameter Estimationįunction dfddm evaluates the density function (or probability density function, PDF) for the Ratcliff diffusion decision model (DDM) using different methods for approximating the full PDF, which contains an infinite sum.Benchmarking Model Fitting to Real-World Data.
Benchmarking the Density Function Approximations.
0 kommentar(er)
