-
Notifications
You must be signed in to change notification settings - Fork 30
/
Copy pathnews.Rmd
executable file
·91 lines (64 loc) · 3.07 KB
/
news.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
---
title: "News"
description: |
Updates and changes
output:
distill::distill_article:
self_contained: true
toc: true
toc_depth: 2
---
```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = FALSE)
```
## 2021-11-11
The backend issue has been resolved and estimates are being produced again.
## 2021-10-21
We are not currently updating estimates due to an issue with our backend infrastructure. We hope to resolve this in the next few weeks.
## 2021-06-28
Stop plotting forecasts. These can still be found in the summary files.
## 2021-01-13
Site design updated to include a page per geography and to reduce page load times.
## 2020-12-17
Due to the ECDC switching to weekly data we have update our global data source to be the WHO.
## 2020-09-18
* Updated methods
## 2020-08-28
* Rebased estimates to use `EpiNow2`.
* Switched data source to `covidregionaldata`.
* Added subnational estimates for Canada.
* Restricted estimates reported to the last 12 weeks of data.
* Updated methods and landing page to link to the new approach.
## 2020-05-17
* Added regional estimates for Brazil and India
* Approximate sampling is now adaptive. Sampling is exact when cases are below 10,000 in a region and after this point it is approximate.
* Incubation period is now explicitly sampled from with right-truncation dealt with in the same way as for imputed onsets.
* Time window for the effective reproduction window is now optimised at every time point.
* Changed to using a generation time estimate rather than a serial interval estimate (deals with the enforced positivity of the serial interval).
* Switched to using the `forecastHybrid` package for generating forecasts as the `fable` package was shown to have a memory leak when used at scale.
* Moved to using a quasipoisson model for estimating the rate of growth and the doubling time.
* Corrected the sources for the UK and German regional data sources.
## 2020-05-02
* Switched to approximate sampling of the delay distribution.
* Forecasts are now only shown for countries/regions that have reported at least 200 cases in the last 7 days.
## 2020-04-25
* Added short term forecasting results.
* Fixed Germany regional map.
* Update plots.
* Fixed doubling time confidence intervals.
* Update palette.
## 2020-04-18
* Updated the UK data source.
* Fixed data extraction for German regions
* Reduced the lower bound on the number of cases needed to be included in the global nowcast from 100 to 60.
* Squashed the NA category from appearing in the USA map.
## 2020-04-09
* Updated all estimates with latest data.
* Corrected a mismatch between the data used for the USA on national and statewide level.
* Added additional text explaining why estimates are lagged.
* Updated wording to highlight date of estiamtes.
## 2020-03-31
* New website to display results.
* Regional breakdowns added for Germany, United Kingdom, and the United States of America.
* National estimates added for all countries that have reported at least 100 cases in a day.
* Update approach to optimise the window used when estimating the Reproduction number.