The MilkBot lactation model is a nonlinear lactation model describing daily milk production
as a function of time(days) since calving (DIM). It is parameterized so that fitted parameter
values are can be interpreted as observable features of the lactation as a whole.
- scale, the overall level of milk production
- ramp, governs the rate of the rise in early lactation
- decay is the rate of exponential decline, most apparent in late lactation
- offset is a small (usually insignificant) correction for time between calving and the theoretical start of lactation
- t is time since calving in days (DIM)
An open-access paper in PeerJ provides a good introduction to the model. Quantifying inter-group variability in lactation curve shape and magnitude with the MilkBot® lactation model
Because it is a nonlinear model, fitting is not straightforward. We have developed and validated a Bayesian fitting engine that estimates fitted parameter values from a set of expected values (the priors) and any number of data points.
This API) describes the API for a planned public service for access to that fitting process.
This API is under development and open for comment and suggestions. When some consensus has been achieved on a workable API, the fitting engine will be ported to this API.