Skip to content

Instantly share code, notes, and snippets.

View yutannihilation's full-sized avatar
🍣
Nobody loves you

Hiroaki Yutani yutannihilation

🍣
Nobody loves you
460 followers · 189 following
View GitHub Profile
@yutannihilation
yutannihilation / reyrender_tokyor.R
Last active September 13, 2020 09:58
library(rayrender)
# generate path data by https://github.com/yutannihilation/rusttype-test
d <- readr::read_csv("/tmp/tmp.csv", col_names = c("x", "y", "id"))
l <- split(d, d$id)
colors <- scales::colour_ramp(c("#FF3499", "#FFECEC"))(seq(from = 0.4, to = 0.0, length.out = length(l)))
dir.create("/tmp/tokyor/")
@yutannihilation
yutannihilation / vector_tile.md
Last active August 17, 2020 00:50
@yutannihilation
yutannihilation / separate_cols.md
Last active July 24, 2020 21:07
separate_cols()

tidyr::separate() cannot handle the cell that contains arbitrary order of values. separate_rows() + pivot_wider() works to some extent, but this cannot handle an empty string or a missing value.

tibble(
  id = 1:4,
  answer = c("a,b", "a,c", "a,b,c", "c")
) %>% 
  # separate() だとできないのでいったん縦向きに広げてから pivot する
  separate_rows(answer) %>%
@yutannihilation
yutannihilation / dev_version_testthat_result.txt
Created July 5, 2020 14:06
Test result of ggplot2 with the dev version of testthat
==> devtools::test()
Loading ggplot2
Testing ggplot2
✓ | OK F W S | Context
✓ | 1 | Adding plot elements
✓ | 17 | test-aes-calculated.r
✓ | 11 | Aesthetics (grouping) [0.4 s]
✓ | 10 | Aes - setting values [0.6 s]
✓ | 45 | Creating aesthetic mappings [1.3 s]
@yutannihilation
yutannihilation / Are_mutate_and_summarise_different.md
Created June 27, 2020 23:22
Are mutate() and summarise() different? 
library(dplyr, warn.conflicts = FALSE)

g <- starwars %>%
  select(name, mass, species) %>%
  group_by(species)

# usual usage of mutate()
d1 <- g %>% 
  mutate(mass_norm = mass / mean(mass, na.rm = TRUE))
@yutannihilation
yutannihilation / tidymodels_xgboost_early_stopping.md
Created May 6, 2020 04:25

Based on the example code on https://www.tidymodels.org/learn/models/parsnip-nnet/

library(tidymodels)
#> ─ Attaching packages ──────────────────────────────────────────── tidymodels 0.1.0 ─
#> ✓ broom     0.5.6          ✓ recipes   0.1.12    
#> ✓ dials     0.0.6          ✓ rsample   0.0.6     
#> ✓ dplyr     0.8.5          ✓ tibble    3.0.1     
#> ✓ ggplot2   3.3.0.9000     ✓ tune      0.1.0     
#> ✓ infer     0.5.1          ✓ workflows 0.1.1
@yutannihilation
yutannihilation / arrow_install.sh
Last active November 22, 2020 06:46
sudo pacman -S libutf8proc
cd cpp
mkdir build
cd build
ARROW_BUILD_TOOLCHAIN=/usr ORC_HOME=/usr DOUBLE_CONVERSION_HOME=/usr cmake \
.. -DARROW_DEPENDENCY_SOURCE=SYSTEM \
-DARROW_PYTHON=ON \
-DCMAKE_BUILD_TYPE=Release \
@yutannihilation
yutannihilation / 和牛.md
Created April 12, 2020 08:38 
library(estatapi)
#> このサービスは、政府統計総合窓口(e-Stat)のAPI機能を使用していますが、サービスの内容は国によって保証されたものではありません。
library(dplyr, warn.conflicts = FALSE)
library(ggplot2)

d <- estat_getStatsData(appId = keyring::key_get("estatapi"), statsDataId = "0003368797")
#> Fetching record 1-2120... (total: 2120 records)

d %&gt;%
@yutannihilation
yutannihilation / different_problems.md
Created April 11, 2020 13:14 
library(readr)

bad_df1 <- readr::read_csv("a,b\n1,2\n-,2", col_types = "nn")
#> Warning: 1 parsing failure.
#> row col expected actual         file
#>   2   a a number      - literal data
bad_df2 <- readr::read_csv("a,b\n1,2\n2,-", col_types = "nn")
#> Warning: 1 parsing failure.
#&gt; row col expected actual file
@yutannihilation
yutannihilation / delika_client.md
Last active April 6, 2020 15:22 
library(httr)

delika_new_jwt <- function() {
  # TODO
  browseURL("https://api.delika.io/v1/auth/token")
}

delika_verb <- function(verb, path, ...) {
  secret <- jsonlite::fromJSON("~/.delika/token.json")