The idea of developing modular software to simulate the dynamics and control of mosquito-borne pathogens originated sometime around 2009 at the Emerging Pathogens Institute, University of Florida. It took much longer than anticipated to finish. The supporting concepts have appeared in various publications. Some algorithms trace back to software packages that never launched. In retrospect, a key challenge was blood feeding, but it was not the only one.
Briefly, ramp.xds and its supporting
satellite packages have replaced MicroMoB, and
exDE. David L Smith developed the prototypes in R, but it
was Sean L. Wu who turned them into the first versions of the
R-packages. In the case of MicroMoB, the software package
was develoepd with the help of Sophie Libkind. The name
exDE was changed first to ramp.xde, which gave
rise to ramp.dts, a discrete-time version of
ramp.xde. After a few months, these two merged and were
rebranded as ramp.xds where development continued. Stable
versions of MicroMoB and exDE were restored to
their last stable versions.
The following is a slightly longer and more detailed timeline:
Late in 2021, MicroMoB
was developed as a software package to build discrete-time dynamical
systems models for the epidemiology, transmission dynamics, and control
of malaria and other mosquito-transmitted pathogens. The software was
born late in 2021. David Smith wrote the first draft of the software.
Sean Wu turned it into a software package and added a lot to it, with
the help of Sophie Libkind.
Late in 2022, exDE
was developed as a software package to build and solve continuous-time
dynamical systems mdoels for the epidemiology, transmission dynamics,
and control of malaria and other mosquito-transmitted pathogens. Once
again, it was David Smith wrote the first draft of the software. Sean Wu
turned it into a software package.
The first version of exDE had been designed to
handle non-autonomous systems, but none of those features had been fully
implemented. During 2023, new features were added. At some point,
exDE was rebranded as ramp.xde. Among the new
features added were:
A major change was implemented in the way ramp.xde
handled delay differential equations. To interalize
computation of terms, a new protocol was developed to add dummy
variables and use lagderivs to internalize computation of
any term passed to any module. New algorithms were added to
Ross-Macdonald models that compute mosquito survival and dispersal
through an incubation period varying in length.
Inits – initial values, had been
handled in a single list, but the protocol was changed so they are now
handled separately, with new functions get_inits() and the
S3 families of functions get_inits_L() and
get_inits_MYZ() and get_inits_X(). A new
function update_inits() resets the initial values to the
last value simulated. The change avoids any issues that migh arise if
two different modules used the same variable name.
Indexing – indices for
variables are now collected in pars$ix
Setup – Several new functions were added to set up basic models.
Solving – The S3 function, now
called xds_solve() was added to solve differential
equations of various classes.
Outputs – The S3 function
parse_deout() was added to parse the output of
deSolve and return the orbits in a set of named lists.
These lists are attached to pars$outputs$orbits. New
functions were also written to compute the dynamical terms that connect
the major components.
Demography – Functionality was added to handle human demographic changes, including births, deaths, migration, and aging.
Hosts – An overhaul was required to handle multiple vertebrate host or vector species.
Forcing – The models have been reconfigured to model malaria as a changing baseline that has been modified by vector control. This required changes to the order of functions that implemented various aspects of forcing by weather and other exogenous variables, and to the return values for functions that computed independent effect sizes
ramp.dts was born in April, 2024.
The goal had always been to bring MicroMoB up to the same
level of functionality as exDE / ramp.xde but
it suddenly seemed easier to clone ramp.xde and replace the
functions
ramp.xds was born when it became
clear that ramp.xde and
ramp.dts had large amounts of overlapping
code.