A Shiny application developed by Jérôme Lavoué on behalf of the Industrial Hygiene team of the Department of Environmental and Occupational Health at the School of Public Health of the Université de Montréal.
Ununoctium (Jean-Mathieu Potvin) collaborated on the development efforts (partial refactoring of the source code) that led to version 4.0.0.
This tool interprets a dataset of exposure measurements (including non detects) with regards to an OEL (Occupational Exposure Limit). In addition to multiple illustrative graphs, it exposes five components.
- Goodness of fit with respect to the lognormal model (graphical evaluation).
- Descriptive statistics.
- Risk assessment based on exceedance of the OEL.
- Risk assessment based on the 95th percentile.
- Risk assessment based on the arithmetic mean.
Calculations are performed using a Bayesian model fit using a Monte Carlo Markov Chain (MCMC) engine. It assumes that the underlying exposure distribution is lognormal.
The underlying Bayesian models and data interpretation procedures are derived from best practices in industrial hygiene data interpretation techniques. They are thoroughly described in Expostats: A Bayesian Toolkit to Aid the Interpretation of Occupational Exposure Measurements (Annals of Work Exposures and Health, Volume 63, Issue 3, April 2019, Pages 267–279).
To serve this application locally, call
.run()This is a shortcut to source file .scripts/entrypoint.R.
This application is deployed to and runs on
shinyapps.io. See file
.scripts/publish-to-shinyapps.R for more information.
4 environments variables are required to push new releases. They must be stored
in a top-level .Renviron file as shown below. This file is ignored by Git
and rsconnect.
# .Renviron
RSCONNECT_ACCOUNT_NAME=lavoue
RSCONNECT_ACCOUNT_TOKEN=<token>
RSCONNECT_ACCOUNT_SECRET=<secret>
APP_VISIBILITY_ON_SHINYAPPS=public
We are actively looking for external collaborators who could help us with supporting more languages. If you are interested, please send an e-mail to [email protected].
Tool 1 relies on package transltr to support multiple languages. Mechanisms are split among two scripts.
- Actions that must be performed before runtime are in
.scripts/find-source-text.R. Notably, this script is in charge of generating the contents of subdirectoryintl/. - Objects required at runtime are defined in
R/intl.R.
Further mechanisms not related to transltr (such as support for ordinal
numbers) are defined in R/helpers.R.
Tool 1 uses placeholders to dynamically insert text into other text. Tokens
such as %s, %i, and %% in the source text must be left as is in
all related translations.
You may submit bugs, request features, and provide feedback by creating an issue on GitHub.