BenjaminWolfe

// background
function timer(ms) { return new Promise(res => setTimeout(res, ms)); }
async function keepGoing(direction, i) {
    await timer(500);
    const directions = {"up": "previous", "down": "next"};
    document.querySelectorAll(`[data-qa="fixed-footer-navigation-${directions[direction]}"]`)[0].click()
    
    await timer(500);
    let focusQ = document.querySelectorAll('[data-qa-focused="true"]')[0]
    let qId = focusQ.getAttribute('data-qa-blockref');

BenjaminWolfe

collect_iteratively <- function(x, size = 500, timeout = 3600) {
  start_time <- Sys.time()
  
  message("pulling ", size, " records at a time...")
  message("starting at ", start_time)
  message("will time out at ", start_time + timeout, " (", timeout, "s later)")
  
  con <- x$src$con
  sql <- dbplyr::db_sql_render(con, x)
  res <- DBI::dbSendQuery(con, sql)

BenjaminWolfe

# hacky but fun code to change the RStudio theme every (e.g.) 15 minutes,
# cycling through the themes in the {rsthemes} package.
# https://gist.github.com/BenjaminWolfe/fc00c480f8a7185779669d2895829069
# 
# start is 9am on the day rstudio::conf(2021L) kicks off.
# applies the current theme, then starts a timer to change it periodically.
# to rotate less often (or more), adjust theme_period below.
# (fractional minutes may not work.)
# note that breaks are just based on lubridate::round_date logic.
# 

BenjaminWolfe

# See `curl.md`. Here is the R equivalent.

library(httr)
library(jsonlite)

make_page <- function(url, headers, data) {
  response <- POST(
    url = url,
    body = toJSON(data, auto_unbox = TRUE),
    config = add_headers(.headers = headers)

BenjaminWolfe

# how to use np.where with multiple columns at once, even whole data frames?

import numpy as np
import pandas as pd

np.random.seed(42)

df1 = pd.DataFrame(np.random.randint(0, 100, size=(100, 4)), columns=list("ABCD"))
df2 = pd.DataFrame(np.random.randint(0, 100, size=(100, 4)), columns=list("ABCD"))
condition = pd.Series(np.random.choice(a=[False, True], size=100, p=[.1, .9]))

BenjaminWolfe

import numpy as np, pandas as pd

df_len = 1000  # integer multiple of 4
np.random.seed(42)

# create a random data frame
df = pd.DataFrame(
    {
        "group_a": np.random.randint(0, df_len / 4, size=df_len),
        "group_b": np.random.randint(0, df_len / 4, size=df_len),

BenjaminWolfe

Say you want to access GitLab.com for work (using your work email address), and you also want to access it for personal reasons (using your personal email address), and you want to do both on the same laptop.

Generate two SSH keys and save them in your ~/.ssh directory. I'll call them work_key and personal_key. As always with SSH keys in GitHub or GitLab, you'll want to generate a public key for each and add it to the corresponding GitLab account.

Then point to both private keys in your ~/.ssh/config file:

BenjaminWolfe

library(tidyverse)
library(repurrrsive) # game of thromes dataset
library(listviewer)  # jsonedit for pretty viewing

jsonedit(got_chars, mode = "view")

really_nested <- tibble(id = 1:30, nested_stuff = got_chars)

really_nested %>%
  hoist(

BenjaminWolfe

#!/bin/bash

declare -A config_link

current_folder=~/files/gladis/datateam/personal/benjamin/config

config_link[authorized_keys]=~/.ssh
config_link[main.code-workspace]=~/files
config_link[.bash_profile]=~
config_link[autoexec.pth]=~/.local/lib/python3.7/site-packages

BenjaminWolfe

# code to be compared to group-and-fill.py
# task: fill specific columns down, within each group, ordering by the order.
library(tidyverse)

df <- tribble(
  ~group, ~order, ~attribute_1, ~attribute_2, ~irrelevant,
  "a",   0,    1,    1,   "hello",
  "a",   2,   NA,    3,    NA    , # this one out of ordr
  "a",   1,    6,   NA,   "world",
  "b",   0,    2,    7,   "foo"  ,