microcovid / microcovid Goto Github PK

View Code? Open in Web Editor NEW

264.0 11.0 54.0 118.46 MB

Estimating the COVID risk of ordinary activities

Home Page: https://www.microCOVID.org

License: MIT License

HTML 0.06% TypeScript 93.17% JavaScript 0.20% Python 5.22% SCSS 0.97% Shell 0.14% Jupyter Notebook 0.23%

covid-19 covid risk-assessment covid19 risk-analysis

microcovid's Introduction

microCOVID.org

Local Development

Requirements:

Node.js 12
Yarn

(See .tool-versions for specific and up-to-date versions).

Install local depenendences
```
$ yarn install
```
Start local server
```
$ yarn start
```
Open http://localhost:3000

Prevalence updates

Install venv to be able to run prevalence updates

$ python3 -m venv .venv
$ source .venv/bin/activate
$ pip install -r requirements-manual.txt

Install the GitHub CLI to be able to using the gh command with the prevalence_helper.sh script
```
# using homebrew (see GitHub page for alternate install methods)
$ brew install gh
```
Get a COVIDActNow.org API key from here. (If you're on the microCOVID team, you can ask for our existing API key.)

Then configure your API key as such:

$ cp .secure-keys.template .secure-keys
# Add your API key to the line in .secure-keys that is listed as CAN_API_KEY

Run a prevalence update. Pick one of the three following examples:

# Run the prevalence update on the auto-prevalence-update branch
$ yarn prevalence
# Run the prevalence update, commit changes, and submit auto-merge pull request
$ yarn prevalence -c
# Run the prevalence update without switching branches (useful when testing changes on a feature branch)
$ yarn prevalence -b
# Specify the COVIDActNow.org API key
$ yarn prevalence -k 111aaa222bbb
# Specify where your virtualenv lives (defaults to `.venv`)
$ yarn prevalence -v .venv-custom

Linting

We use eslint to standardize our style across files. To check your files with the linter, run:

$ yarn lint

If you see warnings, you can attempt to auto-correct them with:

$ yarn fix

If there are errors it can't fix, please fix them manually before committing.

Validating model changes

One of the tests generates computed_scenarios.json, a json of inputs and outputs for various situations. Use model_comparison.ipynb to compare the model behavior before and after model changes.

Updating Community Infection Values

Updating Slider Labels

To update the prevalance slider Labels shown in the community infection Slider component, you can change the property values prevalance_slider_label_1, prevalance_slider_label_2, prevalance_slider_label_3, prevalance_slider_label_4, prevalance_slider_label_5, prevalance_slider_label_6 located in /src/locals/en.json.

Updating Slider Values

You can add 4 values to the slider. To modify these values, you can update the property values prevalance_slider_value_min, prevalance_slider_value_2, prevalance_slider_value_3, prevalance_slider_value_max located in /src/locals/en.json. Please note that values are in multiples of 100. So a value set at 0.8 will correspond to 80 infections. Another example: a value of 7.5 will correspond to 750 infections.

microcovid's People

Contributors

Stargazers

Watchers

Forkers

mikewolfd david-mears richardgill lamgary gyp jankyboy zeno4ever linesarefuzzy cameronmarlow kevinma1515 mattbell raineydavid iwalkinla comradekingu bassettsj robinovitch61 jjst gromadan nelhage jy2wong octern germainz devashleyd justinhaaheim hmrm melonmouse audreybeard ted107mk hazel-nut henryaj tiancheng-luo neurodynamic ghoshsouma nathanawmk taymonbeal bluehat josiah-roberts cmounce ongardie david-allison josephmisiti rtwfroody ruadh egillespie apiology tquigg96 mariehuynh jeanpaulrsoucy bermannoah shawnbiesan2 kyhwana cloudbyteks audunvn nultavenru phrasnolihard

microcovid's Issues

[Blog] Questions should also link to Research Sources

[Blog] Move authors to "contact us"

[Blog] Launch

[Blog] Table widths on mobile

[Blog] Add a mailing list subscription for updates (using catherio's mailchimp)

Audit Presets

Ex: Outdoor dining preset reports 210µCOV in Sacramento, which is much more than we record in the Ibasho spreadsheet.

In general, Number of People Near You seems to be fairly sensitive. We might want to consider giving guidance on this (for instance, how far is "near"?)

Consider performing dataset fetch in frontend

[Blog] can we have a floating table of contents?

http://www.menucool.com/ui/float-panel

Add “low risk” to output

[Blog] Donation at end of conclusion

[Blog] Include “A CIViC DUTY” from https://time.com/5883081/covid-19-transmitted-aerosols

[Blog] Add more images to the post

Regression tests on the results

light exercise as a 3x. heavy exercise as a 6x.

[Blog] FAQ about what if the people are each others' partner

Default to 0.1% annualized budget?

The calculator has a baked-in 1% annualized budget. This makes sense for people in our age group, but may be reckless for:

anyone over the age of 50
anyone whole lives with someone over the age of 50
immunocompromised folks.

I worry we may encourage behavior that puts older folks at risk if we don't clearly warn against using this budget as a blanket rule. Maybe we also should include messaging around being careful about not coming into contact with older folks while carrying >20µCoV

[Blog] "why don't you show this relative to the baseline mortality rate" => this one's tricky, because if young people act as uncautiously as our mortality stats would suggest, on a society-scale we would be contributing to deaths among more-vulnerable people. so we try to mention this, but also tread lightly around it and contextualize it in "risk to other" in the white paper

[Blog] Research Sources gets its own menu item

It is hard to see how to enter specific prevalence data once you have selected a location already

On the calculator, the initial state allows you to enter your own prevalence data, e.g. if you know about your local area and want to use that more accurate data.

However, if you select a location, it is very difficult to see how to then revert your decision, and input your own data. Originally, when making this issue, I believed it was impossible to revert your decision. I had been going to complain that because the form selection is persisted between page reloads, you cannot get around this by reloading the page.

However, I have since found that clicking on 'Select location or enter data' in the first of the two location drop-downs allows me to enter my own data.

It's not just me who has found it difficult to find the way to revert the decision. I shared the site with my partner (while telling her not to share it on further) yesterday, and she commented on this (unsolicited):

it's a shame microcovids doesn't let you have more granularity than 'England'. It would be good if i could enter cases in Hackney or London + population, because i'm guessing the risk that the calculator outputs will be 'diluted' by all the less-risky parts of england

i get that they can't have access themselves to data for everywhere in the world, but it looks like i ought to be able to adjust the pop and case numbers myself, but I can't.

I suggest adding a button saying 'Fill in own data' which has the same effect as clicking 'Select location or enter data'. I most likely will not have time to make this button myself before launch.

[Blog] Add thank you to essential workers in Intermediate Risk

I experience some dissonance reading that healthcare workers are 3x regional, or 8000µCoV in our area, while Sarah's workweek costs 9µCoV.

Since the calculator doesn't support rated PPE, it's not even possible for others to calculate out what their frontline workers risks are.

Perhaps this needs some messaging around "Basic/Intermediate are for people you can't or don't want to have detailed risk discussions with"