Shiny app for decay curve model

UCD-SERG / serocalculator

Estimating Infection Rates from Serological Data

https://ucd-serg.github.io/serocalculator/

4 stars 2 forks source link

Shiny app for decay curve model #72

Closed d-morrison closed 6 months ago

d-morrison commented 8 months ago

We want to build an app that takes the 5 curve parameters as inputs and runs autoplot.curve_params()

[ ] $y_0 < 1000$
[ ] $t_1 < 100$
[ ] label alpha as "(in days)"

kristinawlai commented 8 months ago

Vision: An app that has 5 sliders for y0, y1, t1, alpha, and r for the decay curve so the app will redraw the single curve as a teaching tool. Scaled by min and max for y1 in all of mcmc samples.

kristinawlai commented 8 months ago

Rough sketch:

chrisorwa commented 8 months ago

@d-morrison I need a quick explanation on how the curve is generated. The code has a Markov Chain Monte Carlo file with the y0,y1,t1,alpha and r values which are utilized to generate the elisa units. It takes a while for the code to run and generate the graph. In addition, I'd like to know how the selection of the five values fit into the code.

d-morrison commented 8 months ago

@chrisorwa please take a look at https://github.com/UCD-SERG/serocalculator/pull/80; I've simplified the relevant code a bit. In the shiny app, the user inputs should take the place of the MCMC file; in other words, we should use the inputs to construct a 1-row data.frame and pass that to plot_curve_params_one_ab().

chrisorwa commented 8 months ago

Thanks @d-morrison.

chrisorwa commented 7 months ago

Look into the issue of omitted rows.
Check into log transformation of zero
Change y axis to log scale

kaiemjoy commented 7 months ago

Make increments of y0 smaller from 0,1,2,3,4,5,6
sliders on log scale
check why early growth is not showing, is this an issue of log transforming 0 ??? - for the situation when baseline response is low like 0.5 or 0

chrisorwa commented 7 months ago

Hi @kaiemjoy I have made the changes. However, the early growth is still not showing.