This is the begining of a cheat sheet!

By Johanie Fournier, agr. in rstats viz eda models

October 12, 2023

I can spend a ridiculously big amount of time looking for a certain peace of code that I made few months ago that will solve my current coding problem but can’t remember how to code…

So instead of loosing my time in my archive files or on Google, I decide to start listing here all the snippet of code that I need on hand.

One of this day, I will organize all this into a beautiful cheat sheet, but for now there my little list!

Geospatial

CRS

I need to remember that in the coordinate reference system (CRS) of Quebec:

WGS84:EPSG4326 is a geographic reference system meaning with longitude and latitude values in degree.
NAD83:EPSG3978 is a projected reference system with longitude and latitude values in meter. Better for any kind of calculation

Transform datable to sf object

sf::st_as_sf(.) %>%
  sf::st_transform(., 3978)

sf::st_as_sf(coords = c("longitude", "latitude"))

Add a buffer distance around points

sf::st_buffer(dist=20)

Get coordinate from a sf object

mutate(longitude = sf::st_coordinates(.)[,1],
        latitude = sf::st_coordinates(.)[,2])

Markdown in R

date: "`r format(Sys.time(), '%Y-%m-%d')`"

<script src="https://hypothes.is/embed.js" async></script>

Setup

knitr::opts_chunk$set(include=TRUE, #TRUE = run and include the chunk in document
                      echo   =FALSE, #FALSE = not display the code
                      eval   =TRUE, #FALSE = not run the code in all chunk
                      comment=FALSE,
                      message=FALSE, 
                      warning=FALSE)

Theme Map

theme_map <- function(base_size=9, base_family="") { # 3
	require(grid)
	theme_bw(base_size=base_size, base_family=base_family) %+replace%
		theme(axis.line=element_blank(),
			  axis.text=element_blank(),
			  axis.ticks=element_blank(),
			  axis.title=element_blank(),
			  panel.background=element_blank(),
			  panel.border=element_blank(),
			  panel.grid=element_blank(),
			  panel.spacing=unit(0, "lines"),
			  plot.background=element_blank(),
			  legend.justification = c(0,0),
			  legend.position = c(0,0),
			  plot.title= element_text(size=20, hjust=0, color="black", face="bold"),
		)

emo::ji!

`r emo::ji("collision")`

`r emo::ji("popper")`

`r emo::ji("bomb")`
 
`r emo::ji("bug")`
  
`r emo::ji("chart")`

`r emo::ji("cry")`

`r emo::ji("disaster")`

`r emo::ji("fear")`

`r emo::ji("chart")`

`r emo::ji("flowers")`

`r emo::ji("laugh")`

Parallel Processing

Start

doFuture::registerDoFuture()
n_cores<-parallel::detectCores()-1
future::plan(
  strategy=future::cluster,
  workers=parallel::makeCluster(n_cores)
)

Stop

future::plan(future::sequential)

Tidyverse data manipulation

Replace Inf with NA

df %>% 
  mutate_if(is.numeric, list(~na_if(., Inf))) %>% 
  mutate_if(is.numeric, list(~na_if(., -Inf)))

df %>% 
mutate_at(vars(auc_evi), ~na_if(., Inf))

Rename in a loop

lookup<-c (new_name="old_name")

df %>% 
  rename(any_of(lookup))

Remove cap and accent

df %>% 
mutate(new_name=tolower(stringi::stri_trans_general(old_name, "Latin-ASCII")))

Remove quote in dataframe

df %>%
  mutate(new_colum=gsub("\"", "", old_colum))

Read all rds file in folder

data<-list.files(path="path_to_files", pattern = ".rds", full.names=T) %>%
  map_dfr(readRDS) %>% 
  bind_rows()

Corrrelation Funnel

library(correlationfunnel)

var="name_of_interest_variable"

var_select <-tbl_prep %>% 
  select(var) %>% 
  binarize() %>% 
  select(starts_with(var) & ends_with("_Inf")) %>%
  names()
  
  tbl_prep %>% 
  binarize() %>% 
  correlate(var_select) %>% 
  plot_correlation_funnel() +
  labs(title= paste0(label, " - Correlation Funnel"))

Posted on:: October 12, 2023

Length:: 2 minute read, 370 words

Categories:: rstats viz eda models

Tags:: rstats viz eda models

See Also:: Predicting MO with H2O Models from IRDA data; IRDA soil data; This is a new begining!