ilustat

#DataScience | #rstats | #Statistics

[Dirk Eddelbuettel @eddelbuettel]

Installing packages `sf readr xlsx sqldf tmap DT` with one command in the `r2u` demo container: one command, all binaries, all dependencies, and in seconds (cf https://t.co/iny5H9S573) Works the same way on your Ubuntu server, desktop, laptop cloud instance
#rstats #rspatial

[Dirk Eddelbuettel @eddelbuettel]

R^4 Post #037: Introducing r2u
Two x 19k CRAN binaries for @Ubuntu 22.04 and 20.04
https://t.co/NF8hXzxTUu
#rstats #cran

More details at https://t.co/miK5BmjXu9

[ilustat @ilustat]

@IThinkALot @sharon000 #tidyverse version

library(dplyr)
library(tidyr)
library(ggplot2)
tb <- expand_grid(gp=5:9,time=0:9) %>%
mutate(val=rnorm(n())) %>%
group_by(gp) %>%
nest() %>%
as_tibble() %>%
mutate(plot=lapply(data, function(x) ggplot(x, aes(time, val)) +geom_line()))
tb$plot[3]

[ilustat @ilustat]

@IThinkALot @sharon000 Indeed I ended up repeating each gp (.SD) 9 times. Syntax should be the same as for summary statistics, e.g.,
xx[, .(N = .N, Mean = mean(var), SD = sd(var)), by = .(grp1, grp2)]
I hope my updated response helps! 🙂

[ilustat @ilustat]

@IThinkALot @sharon000 Try:
library(data.table)
library(ggplot2)
dt <- CJ(gp = 5:9, time = 0:9)
dt[, val := rnorm(.N)][
, dt := list(list(.SD)), .(gp)][
, plot := lapply(dt, function(x) ggplot(x, aes(time, val)) + geom_line())]
dt$plot[5]
#rdatatable #ggplot2

[ilustat @ilustat]

@IThinkALot @sharon000 You could replace the ` function(x) ggplot(x, ...` with a custom function.

[ilustat @ilustat]

@IThinkALot @sharon000 Try:
library(data.table)
library(ggplot2)
dt <- CJ(gp = 5:9, time = 0:9)
dt[, val := rnorm(.N)][
, dt := list(list(.SD)), .(gp)][
, plot := lapply(dt, function(x) ggplot(x, aes(time, val)) + geom_line())]
dt$plot[5]
#rdatatable #ggplot2

[ilustat @ilustat]

@nickopotamus @ecojydrology @eddelbuettel Or some of my course notes:

https://t.co/ExHWNvbRss

[ilustat @ilustat]

@nickopotamus @ecojydrology @eddelbuettel Another guide using Women in Parliament data:

https://t.co/dTFEFyYX9B

[Constantino Caetano @Liesareus]

A short thread on the current COVID-19 situation in Portugal.

Portugal reports one of the highest notification Rt in europe (~1.1). Implying an exponential growth of the notification rate. The 14-day cumulative incidence rate is now above the 480/100 000 threshold.

[DS Unplugged Lisbon @dsup_lx]

Portuguese COVID-19 Cases: 7 day rolling mean almost DOUBLES in 8 days. Cases increasing in ALL age groups.

PT COVID-19 Deaths: 7 day rolling mean increasing.

Source: https://t.co/womyuB7P4M
@CEstatisticAUL

[ilustat @ilustat]

This is a nice between @CEstatisticAUL and @ilustat collaboration.

[Centro de Estatística e Aplicações da UL @CEstatisticAUL]

For a quick update on the Portuguese COVID19 data see our 10 day summary table, with a 7 day rolling mean for the new daily cases. Dashboard free ;) #rstats & #rdatatable code in repo - use and share.

https://t.co/bmMWsXhKIj

[Tiago André Marques @TiagoALOMarques]

Cool resource for latest trends in PT #COVID data. Image from the github readme

[ilustat @ilustat]

Thanks @eddelbuettel

[Dirk Eddelbuettel @eddelbuettel]

Thank you @bgreenwell8 -- but credit where credit is due!

This is co-written with @ilustat as a variation on his earlier 'gsir' and best seen side-by-side as comparison.

Choice of tools is a very good thing, never-ending debates notwithstanding.

#rstats

[Centro de Estatística e Aplicações da UL @CEstatisticAUL]

Download now: User friendly version of the Portuguese Daily Covid-19
data (source @DGSaude). Produce plots and summaries with minimal
effort. To get you started example #rstats code given using
#rdatatable and #ggplot2. https://t.co/bmMWsXhKIj

[David Spiegelhalter @d_spiegel]

wow, my article just got a plug by CMO Prof Chris Whitty at the press conference! https://t.co/VQCwoYROK2

[ilustat @ilustat]

@AmirSariaslan @derboyausleu Thanks for suggesting. I have been thinking about it but had to prioritise #SocialDistancing life. I will raise it with @derboyausleu and work on adapting my materials for live streaming.
Any suggestions or links to where it has been done well would be welcome 🙂 #rdatatable

[ilustat @ilustat]

ilustater Saghir will be giving a #rdatatable workshop in Oslo, Norway. Thanks to @derboyausleu for organising it.

[Oslo UseR Meetup @OURmeetup]

Next month, learn how to be fast and tidy https://t.co/ISUx0ug3MN

[ilustat @ilustat]

https://t.co/btjR4gXKvA