about.Rmd

---
title: "About"
output:
  workflowr::wflow_html:
    toc: false
editor_options:
  chunk_output_type: console
---
<style type="text/css">
.center {
  display: block;
  margin-left: auto;
  margin-right: auto;
  width: 65%;
</style>
<style type="text/css">
.centernotshrink {
  display: block;
  margin-left: auto;
  margin-right: auto;
  width: 100%;
}
</style>

## About

This is an effort to provide timely and understandable COVID-19 data visualizations to the public in Orange County, California. The effort is coordinated by the UC Irvine COVID Awareness Group, consisting of students and faculty from the following units at UC Irvine. 


<a href="https://www.stat.uci.edu" target="_blank"><img src = "assets/UCI14_2Line_ICS_Dept_Stats_blue.png" class="center"/></a>

<div class="centernotshrink">
<a href="http://publichealth.uci.edu/ph/_home/" target="_blank"><img src = "assets/UCI14_FLS_PubHealth_Blue.png"/ width="25%"></a>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
<a href="https://cvr.bio.uci.edu" target="_blank"><img src = "assets/UCi17_CtrVirusResearch_blue.png"/ width="25%"></a>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
<a href="https://infectiousdiseaseinitiative.uci.edu" target="_blank"><img src = "assets/U
CI19_Infectious_Disease_Science_Int_2L_blue.png"/ width="30%"></a>
</div>

### UC Irvine COVID Awareness Group Members: 
Damon Bayer, Bernadette Boden-Albala, Isaac Goldstein, Suellen Hopfer, Markelle Kelly, Rachel Longjohn, Vladimir Minin, Andrew Noymer, Daniel M. Parker, Padhraic Smyth.


### Contact information: 
If you would like to know more about technical details and/or to join the group, get in touch with  [Prof. Vladimir Minin](http://vnminin.github.io). For general media inquiries contact [Prof. Suellen Hopfer](https://www.faculty.uci.edu/profile.cfm?faculty_id=6369) and/or [Prof. Andrew Noymer](https://www.faculty.uci.edu/profile.cfm?faculty_id=5373).
If you are a media organization and would like to use our images, you do not need to ask for permission. Just don't forget to acknowledge UC Irvine COVID Awareness Group. 
The content of this site is licensed under the most permissive creative commons license (click on License tab in the navigation bar to learn more). 

## Methodology

### Data
Figures of cases, deaths and hospitalizations over time have been created with publicly available data from the [California Open Data Portal](https://data.ca.gov/group/covid-19). Cases and deaths are counted by date reported, rather than the date the test was administered or date of death (if someone is tested on Friday, and the results are reported the following Tuesday, the case is counted for Tuesday).  

Maps of cases have been created with data provided by the [Orange County Health Care Agency](https://www.ochealthinfo.com).
If an individual tested multiple times, only the first postive test was counted as a case.

### Moving Averages
Our figures present 7 day moving averages. This means that, for cases on July 7th, for example, we report the average number of cases per million people during the period July 1st to July 7th. On July 8th, we report the average number of cases per million people during the period July 2nd to July 8th, etc. We use moving averages as opposed to raw data (such as the actual number of cases per million people on July 7th) because cases can increase or decrease dramatically on a particular day, making it difficult to see trends in the time series of interest. Moving averages help highlight these trends more clearly.

<!-- ### Cases/Deaths/Hospitalizations per Million People -->
In order to calculate, for example, cases per million people, we take the number of cases reported in a county for a day, divide by the 2019 census population for that county and then multiply by 1 million. We do this because different counties have very different population sizes (Los Angeles County has approximately 10 million while Orange County has approximately 3 million) and we want to provide numbers which are comparable across counties.

### Testing Data
The number of daily new cases depends not only on the number of infected individuals in the population, but also on the number of tests performed. 
It would be much better to plot the positivity ratio (positive cases divided by the total number of tests) instead of cases.
Unfortunately, the number of Covid-19 diagnostic tests performed in a day is not available at a county level from the [California Open Data Portal](https://data.ca.gov/group/covid-19) at this time. 
LA and OC counties do have testing data available ([LA COVID-19 data](http://dashboard.publichealth.lacounty.gov/covid19_surveillance_dashboard/), [OC COVID-19 Data](https://data-ocpw.opendata.arcgis.com/datasets/e5ceebe7edba44cc8f875ca54cc2341a?page=18)), but we prefer to use one data source for now.

### Software
We used [R](https://www.r-project.org) software environment, [tidyverse](https://www.tidyverse.org) collection of R packages, [workflowr](https://jdblischak.github.io/workflowr/) R package, and [QGIS](https://qgis.org/en/site/) geographic information system software.