davidclarance

def poisson_bootstrap_nth_percentile_and_ci(
    data: List[float], n_iter: int = 1000, alpha: float = 0.95, percentile: float = 95
) -> Tuple[float, float, float, float]:

    data = np.array(data)
    n = len(data)

    weights = np.random.poisson(1, (n_iter, n))

    percentiles = []

davidclarance

def bootstrap_nth_percentile_and_ci(
    data: List[float], n_iter: int = 1000, alpha: float = 0.95, percentile: float = 95
) -> Tuple[float, float, float, float]:
    """Compute the nth percentile and confidence interval using bootstrapping

    Args:
        data (List[float]): List of data
        n_iter (int, optional): Number of bootstrap samples. Defaults to 1000.
        alpha (float, optional): Confidence level. Defaults to 0.95.
        percentile (float, optional): Percentile to compute. Defaults to 95.

davidclarance

def bootstrap_mean_and_ci(
    data: List[float], n_iter: int = 1000, alpha: float = 0.95
) -> Tuple[float, float, float]:
    n = len(data)

    bootstrap_samples = np.random.choice(data, (n_iter, n), replace=True)

    means = bootstrap_samples.mean(axis=1)

    mean = means.mean()

davidclarance

WITH customers AS (
  SELECT 'Customer1' AS customer_id UNION ALL
  SELECT 'Customer2' AS customer_id UNION ALL
  SELECT 'Customer3' AS customer_id UNION ALL
  SELECT 'Customer4' AS customer_id
)
SELECT 
  c.customer_id, 
  CONVERT(BIGINT, SUBSTRING(HASHBYTES('MD5', c.customer_id), 1, 5)) % 100 AS assigned_bucket	
FROM customers c;

davidclarance

CREATE TABLE IF NOT EXISTS customers AS
SELECT 'Customer1' AS customer_id UNION ALL
SELECT 'Customer2' AS customer_id UNION ALL
SELECT 'Customer3' AS customer_id UNION ALL
SELECT 'Customer4' AS customer_id;


SELECT 
    c.customer_id,
    ('0x' || md5(c.customer_id)[:10])::int64 % 100 as assigned_bucket

davidclarance

import hashlib

probability_assignments = {"Control": 50, "Variant 1": 30, "Variant 2": 20}

random_customer_ids = ["Customer1", "Customer2", "Customer3", "Customer4"]


def get_hash(customer_id):
    hash_object = hashlib.md5(customer_id.encode())
    return hash_object.hexdigest()[:10]

davidclarance

# Function adapted from https://github.com/AfricaBirdData/ABAP/
# Only pulls presence data

library("tidyverse")
library("purrr")

jsonToTibble <- function(jsonfile) {
  out <- jsonfile %>%
    lapply(function(x) {
      x[sapply(x, is.null)] <- NA

davidclarance

# 0. set up ---------------------------------------------------------------

# libraries
library(rabm)
library(tidyverse)
library(geojsonio)
library(broom)
library(ggplot2)
library(gganimate)

davidclarance

library(tidyverse)

files <- list.files(path = "./data/", pattern = "*.csv")

df <- files %>% 
    map(function(x) {
        read_csv(paste0("./data/", x))
    }) %>%
    reduce(rbind)