jrosell’s gists

jrosell / ai-evals.R

Created July 2, 2025 16:36

Do you know that you can evaluate IA models and compare their performance? Here an example using the {vitals} package from @posit_pbc by @simonpcouch Evals, evals, evals

	rlang::check_installed(c("vitals", "ellmer", "dplyr", "ggplot2"))

	library(vitals)
	library(ellmer)
	library(dplyr)
	library(ggplot2)

	eval_df <- tibble(
	input = c("What's 2+2?", "What's 2+3?", "What's 2+4?"),
	target = c("4", "5", "6")

jrosell / formula_typed_functions.R

Last active July 1, 2025 22:22

An experiment creating new functions with type validation usinf formulas

	# An experiment creating new functions with type validation using formula notation in R.
	# For example, for two double argugment inputs and integer output the formula would be: integer ~ double + double

	library(rlang)
	library(testthat)

	type_check_fn <- function(type_name) {
	switch(as.character(type_name),
	"integer" = is.integer,
	"double" = is.double,

jrosell / run_setup.qmd

Last active June 20, 2025 17:13

Reset the environment variables and execute the setup chunk from other R chunks in quarto. USE WITH CAUTION AT YOUR OWN RISK

	rm(list = ls()); rstudioapi::getActiveDocumentContext()$path \|>
	parsermd::parse_rmd(parse_yaml = FALSE) \|>
	parsermd::rmd_select("setup") \|>
	parsermd::as_document() \|>
	purrr::keep(\(x) !stringr::str_detect(x, '^```\|^#')) \|>
	parse(text = _) \|>
	eval()

jrosell / google-ads-2-google-sheets.js

Created June 16, 2025 12:49

Script in Google Ads account to export data to a Google Sheets

	function main() {
	var spreadsheetUrl = 'https://docs.google.com/spreadsheets/d/id/edit?usp=sharing';
	var spreadsheet = SpreadsheetApp.openByUrl(spreadsheetUrl);

	var pivotedSheet = getOrCreateSheet(spreadsheet, 'PivotedConversionActions'); // New pivoted data sheet

	var firstDate = '2022-01-01';
	var startDate = getDateNDaysAgo(200);
	var endDate = getDateNDaysAgo(1);

jrosell / plumber2-api-json.R

Created June 6, 2025 12:54

	rlang::check_installed(c("yyjsonr", "plumber2", "httr2", "bench", "callr", "palmerpenguins", "ggbeeswarm", "ggplot2"))

	library(plumber2)
	library(palmerpenguins)
	data(package = 'palmerpenguins')

	get_penguins <- \(n = 100) {
	np <- nrow(penguins)
	idx <- sample(1:np, n, replace = TRUE)
	penguins[idx, ]

jrosell / dplyr-ggplot2-shiny-barplot.R

Last active June 5, 2025 16:46

	library(shiny)
	library(dplyr)
	library(ggplot2)
	library(bslib)

	set.seed(42)
	data <- data.frame(
	category = c("A", "B", "C", "D", "E"),
	value = sample(10:100, 5)
	)

jrosell / api.domain.com.conf

Last active March 27, 2025 17:41

Example of a JavaScript fetching country code from API. I used a apache2 and ubuntu service for an R API using plumber and ipinfo.io service.

	<VirtualHost *:80>
	ServerAdmin [email protected]
	ServerName api.domain.com
	ProxyPass / http://127.0.0.1:8000/
	ProxyPassReverse / http://127.0.0.1:8000/
	</VirtualHost>

jrosell / country_from_coordinates_api.R

Created March 27, 2025 12:26

An example service using R API to obtain the country from latitude and longitude coordinates.

	# plumber::plumb("country_from_coordinates_api.R")$run(port = 8000)

	library(plumber)
	library(httr2)
	library(rlang)

	#* @serializer html
	#* @get /
	function() {
	html_content <- '<!DOCTYPE html>

jrosell / clickstream_data_analysis.R

Last active March 27, 2025 09:58

An example of data model to perform clickstream data analysis in R for an ecommerce website or app.

	``` r
	suppressPackageStartupMessages({
	library(dplyr)
	library(tidyr)
	})



	click_data <- tibble(
	click_id = c("u1", "u1", "u1", "u1", "u1", "u1", "u1", "u1", "u1", "u2", "u2", "u2", "u3", "u3", "u2"),

jrosell / install_and_load.R

Last active June 12, 2025 15:28

This is how I like to check for required installed versions and load packages in my R scripts, including non-CRAN packages. It requires {rlang} and {pak} packages.

	if (!requireNamespace('rlang')) stop("Please, run install.packages('rlang')")
	rlang::check_required("pak")
	pkgs <- rlang::chr(
	rlang = "r-lib/[email protected]",
	purrr = "purrr",
	tidyverse = "tidyverse/tidyverse",
	tidymodels = "tidymodels/tidymodels",
	jrrosell = "jrosell/jrrosell@main",
	)
	pak::pak(pkgs)

	rm(list = ls()); rstudioapi::getActiveDocumentContext()$path \|>
	parsermd::parse_rmd(parse_yaml = FALSE) \|>
	parsermd::rmd_select("setup") \|>
	parsermd::as_document() \|>
	purrr::keep(\(x) !stringr::str_detect(x, '^```\|^#')) \|>
	parse(text = _) \|>
	eval()

Jordi Rosell jrosell