aboutsummaryrefslogblamecommitdiff
path: root/vignettes/web_only/saem_benchmarks.rmd
blob: ae61f1534ca005fee429b75fc227566879cf7837 (plain) (tree)
1
2
3
4


                                         
                                                           




























                                                                               
                                                      







                                                                                                                                   






                                                                                          



























































































































                                                                                       
                                                         







































                                                                           
---
title: "Benchmark timings for saem.mmkin"
author: "Johannes Ranke"
date: Last change 17 February 2023 (rebuilt `r Sys.Date()`)
output:
  html_document:
    toc: true
    toc_float: true
    code_folding: show
    fig_retina: null
vignette: >
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r, include = FALSE}
library("knitr") # For the kable() function
opts_chunk$set(tidy = FALSE, cache = FALSE)
library("mkin")
```

Each system is characterized by operating system type, CPU type,
mkin version, saemix version and R version. A compiler was available, so if no
analytical solution was available, compiled ODE models are used.

Every fit is only performed once, so the accuracy of the benchmarks is limited.

```{r include = FALSE}
cpu_model <- benchmarkme::get_cpu()$model_name
# Abbreviate CPU identifiers
cpu_model <- gsub("AMD ", "", cpu_model)
cpu_model <- gsub("Intel\\(R\\) Core\\(TM\\) ", "", cpu_model)
cpu_model <- gsub(" Eight-Core Processor", "", cpu_model)
cpu_model <- gsub(" 16-Core Processor", "", cpu_model)
cpu_model <- gsub(" CPU @ 2.50GHz", "", cpu_model)

operating_system <- Sys.info()[["sysname"]]
mkin_version <- as.character(packageVersion("mkin"))
saemix_version <- as.character(packageVersion("saemix"))
R_version <- paste0(R.version$major, ".", R.version$minor)
system_string <- paste0(operating_system, ", ", cpu_model, ", mkin ", mkin_version, ", saemix ", saemix_version, ", R ", R_version)



if (dir.exists("~/git/mkin")) {
  benchmark_path = normalizePath("~/git/mkin/vignettes/web_only/saem_benchmarks.rda")
} else {
  benchmark_path = normalizePath("~/projects/mkin/vignettes/web_only/saem_benchmarks.rda")
}
load(benchmark_path)

# Initialization 14 November 2022
#saem_benchmarks <- data.frame()

saem_benchmarks[system_string, c("CPU", "OS", "mkin", "saemix", "R")] <-
  c(cpu_model, operating_system, mkin_version, saemix_version, R_version)
```

For the initial mmkin fits, we use all available cores.

```{r setup}
n_cores <- parallel::detectCores()
```

## Test data

Please refer to the vignette `dimethenamid_2018` for an explanation
of the following preprocessing.

```{r dimethenamid_data}
dmta_ds <- lapply(1:7, function(i) {
  ds_i <- dimethenamid_2018$ds[[i]]$data
  ds_i[ds_i$name == "DMTAP", "name"] <-  "DMTA"
  ds_i$time <- ds_i$time * dimethenamid_2018$f_time_norm[i]
  ds_i
})
names(dmta_ds) <- sapply(dimethenamid_2018$ds, function(ds) ds$title)
dmta_ds[["Elliot"]] <- rbind(dmta_ds[["Elliot 1"]], dmta_ds[["Elliot 2"]])
dmta_ds[["Elliot 1"]] <- NULL
dmta_ds[["Elliot 2"]] <- NULL
```

## Test cases

### Parent only

```{r parent_only}
parent_mods <- c("SFO", "DFOP", "SFORB", "HS")
parent_sep_const <- mmkin(parent_mods, dmta_ds, quiet = TRUE, cores = n_cores)
parent_sep_tc <- update(parent_sep_const, error_model = "tc")

t1 <- system.time(sfo_const <- saem(parent_sep_const["SFO", ]))[["elapsed"]]
t2 <- system.time(dfop_const <- saem(parent_sep_const["DFOP", ]))[["elapsed"]]
t3 <- system.time(sforb_const <- saem(parent_sep_const["SFORB", ]))[["elapsed"]]
t4 <- system.time(hs_const <- saem(parent_sep_const["HS", ]))[["elapsed"]]
t5 <- system.time(sfo_tc <- saem(parent_sep_tc["SFO", ]))[["elapsed"]]
t6 <- system.time(dfop_tc <- saem(parent_sep_tc["DFOP", ]))[["elapsed"]]
t7 <- system.time(sforb_tc <- saem(parent_sep_tc["SFORB", ]))[["elapsed"]]
t8 <- system.time(hs_tc <- saem(parent_sep_tc["HS", ]))[["elapsed"]]
```

```{r}
anova(
  sfo_const, dfop_const, sforb_const, hs_const,
  sfo_tc, dfop_tc, sforb_tc, hs_tc) |> kable(, digits = 1)
```

The above model comparison suggests to use the SFORB model with two-component
error. For comparison, we keep the DFOP model with two-component error,
as it competes with SFORB for biphasic curves.

```{r}
illparms(dfop_tc)
illparms(sforb_tc)
```

For these two models, random effects for the transformed parameters `k2` and
`k_DMTA_bound_free` could not be quantified.


### One metabolite

We remove parameters that were found to be ill-defined in the parent only fits.

```{r one_metabolite, message = FALSE}
one_met_mods <- list(
  DFOP_SFO = mkinmod(
    DMTA = mkinsub("DFOP", "M23"),
    M23 = mkinsub("SFO")),
  SFORB_SFO = mkinmod(
    DMTA = mkinsub("SFORB", "M23"),
    M23 = mkinsub("SFO")))

one_met_sep_const <- mmkin(one_met_mods, dmta_ds, error_model = "const",
  cores = n_cores, quiet = TRUE)
one_met_sep_tc <- mmkin(one_met_mods, dmta_ds, error_model = "tc",
  cores = n_cores, quiet = TRUE)

t9 <- system.time(dfop_sfo_tc <- saem(one_met_sep_tc["DFOP_SFO", ],
    no_random_effect = "log_k2"))[["elapsed"]]
t10 <- system.time(sforb_sfo_tc <- saem(one_met_sep_tc["SFORB_SFO", ],
    no_random_effect = "log_k_DMTA_bound_free"))[["elapsed"]]
```

### Three metabolites

For the case of three metabolites, we only keep the SFORB model in order
to limit the time for compiling this vignette, and as fitting in parallel
may disturb the benchmark. Again, we do not include random effects that
were ill-defined in previous fits of subsets of the degradation model.

```{r}
illparms(sforb_sfo_tc)
```

```{r three_metabolites, message = FALSE}
three_met_mods <- list(
  SFORB_SFO3_plus = mkinmod(
    DMTA = mkinsub("SFORB", c("M23", "M27", "M31")),
    M23 = mkinsub("SFO"),
    M27 = mkinsub("SFO"),
    M31 = mkinsub("SFO", "M27", sink = FALSE)))

three_met_sep_tc <- mmkin(three_met_mods, dmta_ds, error_model = "tc",
  cores = n_cores, quiet = TRUE)

t11 <- system.time(sforb_sfo3_plus_const <- saem(three_met_sep_tc["SFORB_SFO3_plus", ],
    no_random_effect = "log_k_DMTA_bound_free"))[["elapsed"]]
```

```{r results, include = FALSE}
saem_benchmarks[system_string, paste0("t", 1:11)] <-
  c(t1, t2, t3, t4, t5, t6, t7, t8, t9, t10, t11)
save(saem_benchmarks, file = benchmark_path, version = 2)
# Hide rownames from kable for results section
rownames(saem_benchmarks) <- NULL
```

## Results

Benchmarks for all available error models are shown. They are intended for
improving mkin, not for comparing CPUs or operating systems. All trademarks
belong to their respective owners.

### Parent only

Constant variance for SFO, DFOP, SFORB and HS.

```{r, echo = FALSE}
kable(saem_benchmarks[, c(1:4, 6:9)])
```

Two-component error fits for SFO, DFOP, SFORB and HS.

```{r, echo = FALSE}
kable(saem_benchmarks[, c(1:4, 10:13)])
```

### One metabolite

Two-component error for DFOP-SFO and SFORB-SFO.

```{r, echo = FALSE}
kable(saem_benchmarks[, c(1:4, 14:15)])
```

### Three metabolites

Two-component error for SFORB-SFO3-plus

```{r, echo = FALSE}
kable(saem_benchmarks[, c(1:4, 16)])
```

Contact - Imprint