kathsherratt

# Run full evaluation sequence for all forecast models saved in data-processed/

# Sources for this code taken from:
# https://github.com/epiforecasts/covid19-forecast-hub-europe/blob/main/code/reports/compile-evaluation-reports.r
# https://github.com/epiforecasts/covid19-forecast-hub-europe/blob/main/code/reports/evaluation/evaluation-report.Rmd
# https://github.com/epiforecasts/EuroForecastHub/blob/main/R/score_models.R
# https://github.com/epiforecasts/covid19-forecast-hub-europe/tree/main/code/reports/rmdchunks

# Set up ------------------------------------------------------------------
# R session working directory should be set to a local clone of covid19-forecast-hub-europe/

kathsherratt

library(dplyr)
library(covidHubUtils)
library(ggplot2)

# Set up
min_date <- Sys.Date() - 6 * 7 # set an earliest date to display (eg six weeks)
location_names <- hub_locations_ecdc$location_name # set which countries to display (eg all countries with data)

# Get data
ecdc_daily <- load_truth(truth_source = "ECDC", 

kathsherratt

# test forecast file

library(dplyr)
library(tibble)
library(readr)
library(purrr)
library(here)

# `tests` contains 9 tibbles each to test one element of validation
# Each tibble is a forecast with one valid row (location A) and one invalid row (location B)

kathsherratt

# Plot Euro hub ensemble forecasts by European region

library(here)
library(covidHubUtils)
library(countrycode)
library(dplyr)
library(lubridate)
library(ggplot2)
library(tidyr)
library(purrr)

kathsherratt

# Count teams contributing to European Forecast hub over time
library(gh)
library(purrr)
library(dplyr)
library(lubridate)
library(ggplot2)
library(ggrepel)

# Get model names / dates from github ---------------------------
model_names <- gh(paste0("/repos/covid19-forecast-hub-europe/covid19-",

kathsherratt

# Create a template scenario projection
library(dplyr)
library(tidyr)
library(readr)

# Variables
origin_date <- as.Date("2022-05-22")
scenario_id <- paste0(c("A", "B", "C", "D"), "-", origin_date)
age_group <- c("0-15", "16-64", "65-130", "0-130")
target_variables <- c("inc case", "inc hosp", "inc death", "inc inf")

kathsherratt

# Source: code by Hugo Gruson (Github @Bisaloo)
# https://github.com/covid19-forecast-hub-europe/covid19-forecast-hub-europe/blob/main/.github/workflows/create-parquet.yml

library(readr)
library(arrow)
library(dplyr)
library(lubridate)
locations <- read_csv(
  here::here("data-locations", "locations_eu.csv")) |>
  dplyr::select(location, location_name, population)

kathsherratt

# Gets variants of concern dataset from ECDC
# For each variant/country finds the first period
#  with variant % greater than 5% and up to 50%
#  before the first peak for that variant
# Note ECDC data are in weeks, dates are indexed to Saturdays to match target end dates
#
# Example
#   variant_names <- c("B.1.617.2" = "Delta", "B.1.1.529" = "Omicron")
#   forecast_variants <- download_variant_introduction(introduction_percent = 5,
#                                                   country_names = "Germany",

kathsherratt

---
title: "Appendix: Responses to improving conditions for outbreak modelling"
date: "2024"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = FALSE, message = FALSE, warning = FALSE)
library(knitr)
library(googlesheets4)