noamross

Research Software Engineer, Infrastructure - EcoHealth Alliance

EcoHealth Alliance seeks a research software engineer to build and maintain technical infrastructure to support scientific research across our field, lab, and computational teams. We are seeking someone with the ability to learn and use a variety of system administration and development operations tools, and to work with a diverse, multi-disciplinary team that includes scientists from across the globe. This is a key position at a rapidly growing NGO with diverse funded scientific research programs around the world focused on conservation

noamross

QUANTITATIVE ECOLOGY RESEARCH SCIENTIST

https://www.ecohealthalliance.org/career/quantitative-ecology-research-scientist

EcoHealth Alliance is seeking a quantitative ecologist to work closely with our team on multiple projects studying bat-borne zoonotic viruses. We are seeking someone with excellent statistical and computational skills, strong communication skills, and the ability to work with a diverse, multi-disciplinary team that includes scientists from across the globe. This is a key position at a rapidly growing NGO with diversely funded scientific research programs around the world focused on conservation and zoonotic disease emergence. This position requires experience with, and an ability to learn, a variety of ecological and epidemiological statistical techniques including generalized linear and additive, spatiotemporal, mixed-effect, Bayesian mixture and occupancy models, as well as proficiency with R and reproducible methods. Responsibilities include leading analysis of longitudinal and spatial

noamross

library(mgcv)
library(spdep)
library(sf)
library(ggplot2)

# Columbus crime data from mgcv/spdep
example(columbus)

# Make a neighborhood object compatible with mgcv's MRF
nb <- poly2nb(columbus)

noamross

library(ggplot2)

dat <- data.frame(a = letters[1:10], b = exp(rnorm(10)))
ggplot(dat, aes(x = a, y = b)) + geom_col() + scale_y_log10() # Basic log scale transform
trans = scales::trans_new("logshift", \(x) log10(x*10), \(x) (10^x)/10) # Custom transformation
ggplot(dat, aes(x = a, y = b)) + geom_col() + scale_y_continuous(trans = trans) # Use custom transformation
ggplot(dat, aes(x = a, y = b)) + geom_col() + scale_y_continuous(trans = trans, breaks = c(0.25, 0.5, 1, 2, 4, 8)) # Add breaks

ggplot(dat, aes(x = a, y = b)) + 
  geom_col() + 

noamross

library(sys)
library(RMariaDB)
library(DBI)
library(withr)

# Clean up from previous session
unlink("doltdb", recursive = TRUE)
try(dbDisconnect(conn), silent = TRUE)

noamross

library(leaflet)
library(tidyverse)

outline <- quakes[chull(quakes$long, quakes$lat),]

map <- leaflet(quakes) %>%
  addTiles(group = "OSM (default)") %>%
  # Overlay groups
  addCircles(~long, ~lat, ~10^mag/5, stroke = F, group = "Quakes") %>%
  addPolygons(data = outline, lng = ~long, lat = ~lat,

noamross

library(ggplot)
library(sf)
library(ggspatial)
library(snapbox)
library(raster)

worldclim <- raster::getData('worldclim', var='prec', res=2.5, lon = 142.61, lat = -23.64)

area <- st_bbox(
  c(xmin = 101.03, ymin = -44.82, xmax = 168.23, ymax = -9.09),

noamross

#!/bin/bash

funzip $1|  # uncompress first file in zip
tr -d '\000' | #remove null characters
sed "/^\s*$/d; s/ \{1,\}\t/\t/g; s/\t \{1,\}/\t/g; s/\r//" |  #removes empty lines, whitespace around tabs, extra newlines
cut -s -f 1,3,4,5,6,8,12,13,14,15,16,17,18,19,20,21,23,24,25,26,34,35,36,38,40,42,44,45,46,85,86,87,88,89 #| #only select certain columns
pv -N Process -c  |
gzip -9 | 
pv -N Compress -c > $1.gz

noamross

#' Create a random effect basis with integers rather than factors
#' @import mgcv
#' @export
smooth.construct.re2.smooth.spec <- function (object, data, knots) {
  if (!is.null(object$id))
    stop("random effects don't work with ids.")
  if(any(sapply(data, is.numeric))) data <- lapply(data, as.factor)   ## <-- All I did was this (and below)
  form <- as.formula(paste("~", paste(object$term, collapse = ":"),
                           "-1"))
  object$X <- model.matrix(form, data)

noamross

#' Function to partition unequal sized groups into shards of similar size.
#'
#' Based on the greedy algorithm described in [The Wikipedia article on the
#' the partitioning problem](https://en.wikipedia.org/wiki/Partition_problem#The_greedy_algorithm)
#'
#' @param groups_vector An integer vector of group ids, such as a group ID
#'   column in a data frame
#' @param n_shards The number of shards to split groups up into
#' @examples
#' n_groups <- 200