https://github.com/emitanaka/sizzled

An R-package to create experiments that require sample size calculation
https://github.com/emitanaka/sizzled

r-package

Last synced: 9 days ago
JSON representation

An R-package to create experiments that require sample size calculation

• Host: GitHub
• URL: https://github.com/emitanaka/sizzled
• Owner: emitanaka
• Created: 2021-01-02T22:55:11.000Z (almost 4 years ago)
• Default Branch: main
• Last Pushed: 2021-07-02T09:33:00.000Z (over 3 years ago)
• Last Synced: 2024-08-06T03:03:57.083Z (3 months ago)
• Topics: r-package
• Language: R
• Homepage: https://sizzled.emitanaka.org
• Size: 157 KB
• Stars: 10
• Watchers: 2
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.Rmd

Awesome Lists containing this project

README

        
---

output: github_document

---

```{r, include = FALSE}

knitr::opts_chunk$set(

  collapse = TRUE,

  comment = "#>",

  fig.path = "man/figures/README-",

  out.width = "100%"

)

```

# sizzled 

[![Lifecycle: experimental](https://img.shields.io/badge/lifecycle-experimental-orange.svg)](https://www.tidyverse.org/lifecycle/#experimental)

The aim of sizzled package is to use some foundational aspects of edibble package but include extra functionality to support experiments that initially have an unknown structure, i.e. requiring calculation for sample size for some units. 

(WIP) There is not much to see here.

Below is not implemented. 

```{r, eval = FALSE}

library(edibble)

library(sizzled)

out <- start_design() %>% 

  set_trts(treat = c("placebo", "vaccine")) %>% 

  set_units(person = symbol("npeople")) %>% 

  allocate_trts(treat ~ person) %>% 

  randomise_trts() %>% 

  serve_table()

# trying different sample sizes, 

# output here is a list

out %>% 

  set_symbols(npeople = c(40, 100, 200))

# standard sample size calculator

# non-vectorised version

out %>% 

  set_symbols(npeople = siz(power = 80,

                            sig_level = 0.05,

                            alternative = "two.sided",

                            effect = 3,

                            target_group = treat))

# vectorised version

out %>% 

  set_symbols(npeople = sizz(...))

# proportion version

out %>% 

  set_symbols(npeople = sizp(...))

# one that relies on other unit structures, etc

out %>% 

  set_symbols(npeople = sizzz(...))

# add more z to have other variations, but more z usually means more

# complex structure

```

One shouldn't sizzle too much lest you overcook your edibble design. 

The sizzling sound also could be the sound of the statistician silently boiling in anger when told to do yet another sample size calculation from their non-statistical colleague. Beware of that sizzling statistician.