Skip to content

Latest commit

 

History

History
300 lines (171 loc) · 7.78 KB

NEWS.md

File metadata and controls

300 lines (171 loc) · 7.78 KB

squire 0.7.0

  • The drjacoby package can now be used to fit mcmc for deterministic model
  • drjacoby_mcmc function analogous to pmcmc for fitting models
  • convert_log_likelihood_func_for_drjacoby helps convert log likelihood functions, e.g. calc_loglikelihood into suitable likelihood functions that can be correctly passed as log_likelihood to drjacoby_mcmc.
  • convert_log_prior_func_for_drjacoby does the same for priors
  • The above changes are because drjacoby requires initial parameters and ranges to be passed as a data.frame internally, which means dates are internally handled as numerics, e.g. "2020-01-01" is 18262. Consequently, the sampled parameters from drjacoby_mcmc will have start_date values as numerics. See sample_drjacoby function for how these are handled correctly to align with converted log likelihood function draws.

squire 0.6.11

  • parse_durations bug fix. Handles dur_E, dur_IMild and dur_ICase correctly now

squire 0.6.10

  • sample_pmcmc bug fix dor negative posterior checks incorrectly.

squire 0.6.9

  • atol and rtol can be passed to pmcmc fitting as named arguments in pars_obs to be used in run_deterministic_comparison
  • pmcmc can now be used to fit to PCR prevalence data (deterministic model only) by passing pcr_df and pcr_det as named list elements of pars_obs

squire 0.6.8

  • run_deterministic_comparison patches for sero fitting and argument checks

squire 0.6.7

  • pmcmc can now be used to fit to serology data (deterministic model only) by passing sero_df and sero_det as names list elements of pars_obs

squire 0.6.6

  • projections can be used now for nimue models

squire 0.6.5

  • pmcmc now accepts dates of change for vaccine efficacy for nimue

squire 0.6.4

  • pmcmc changes to adapt for nimue class models and vaccines

squire 0.6.3

  • beta_est accepts nimue models

squire 0.6.2

  • Typo in dur_not_get_mv_survive

squire 0.6.1

  • pmcmc(..., scaling_factor) allows scaling factors to be passed to pmcmc
  • Correction to deterministic_model model class

squire 0.6.0

  • Waning added to squire

squire 0.5.8

  • Updated to use odin 1.1.0
  • odin_is_discrete new unexported function

squire 0.5.7

  • Github actions included for automatic release tagging

squire 0.5.6

  • pmcmc(..., gibbs_sampling = TRUE) implements Gibbs sampler

squire 0.5.5

  • population$n are all integer class

squire 0.5.4

  • beta_est accepts apothecary models
  • Updated json odin files that weren't compiled correctly

squire 0.5.3

  • default_durations contains all durations
  • default_durations and default_probs now exported for ease
  • run_deterministic_SEIR_model has odin pars exported now

squire 0.5.2

  • Parameters vignette updated to include new parameters
  • run_deterministic_comparison bug fix for when transmission is eliminated

squire 0.5.1

  • projections bug fixed for setting time varying durations
  • Taiwan healthcare capacity added
  • default_durations updated for incorrect hospital durations

squire 0.5.0

0.5.0 Major

  • Parameter Updates. See parameters vignette for update.
  • triggering function for doing projections with model state triggered R0 changes
  • Time-varying durations of ICU and hospital stay

0.5.0 Minor

  • projections allows for different args to be passed to each replicate
  • projections allows for additional user model args to be passed

squire 0.4.34

  • pmcmc argument treated_deaths_only will use deaths from treatment in likelihood

squire 0.4.33

  • format_output accepts deaths_treatment and deaths_no_treatment

squire 0.4.32

  • intervention_dates_for_odin fills dates earlier than date_R0_change

squire 0.4.31

  • Correction to random walk parameters to more accurately continue the walk

squire 0.4.30

  • evaluate_Rt_pmcmc takes Rt_rw_n for random walk esque shifts

squire 0.4.29

  • format_output backwards compatible for incidence

squire 0.4.28

  • calibrate patched as dt argument was not being passed through
  • format_output corrected for hospital/ICU incidence

squire 0.4.27

  • evaluate_Rt_pmcmc patch for non specified Rt_shift in 3 parameter version

squire 0.4.26

  • pmcmc more flexible for varying which parameters are fitted.
  • evaluate_Rt_pmcmc altered for increased flexibility

squire 0.4.25

  • format_output now returns hospital_incidence and ICU_incidence
  • evaluate_Rt_pmcmc patch for Meff_switch_date earlier than date_R0_change[1]

squire 0.4.24

  • Meff_pl can be scaled or not in relation to Meff

squire 0.4.23

  • pmcmc no longer takes Rt_func as hard coded given Meff_pl complexity

squire 0.4.22

  • Johnstone Change approach implemented for pmcmc RM

squire 0.4.21

  • evaluate_Rt patch for roll of Meff

squire 0.4.20

  • evaluate_Rt allows for date_Meff_change to be later/earlier than date_R0_change ends

squire 0.4.19

  • Large overhaul to make projections and plotting work with deterministic models

squire 0.4.18

  • calc_loglikelihood fix for reduced date vectors

squire 0.4.17

  • pars_init for pmcmc can be provided as a list for where each chain starts

squire 0.4.16

  • pmcmc method for 4 parameter estimation.

squire 0.4.15

  • projections now matches by column names when including projection outputs
  • Patch to t0_variables for new Rt_func approach

squire 0.4.14

  • replace chained binomials with mutivariate hypergeometric sampling in stochastic model
  • stochastic model to output number newly requiring ICU beds at each timepoint

squire 0.4.13

  • calibrate now takes an Rt_func argument for specifying Rt calculation

squire 0.4.12

  • Explicit Meff inclusion in R0_date_particle_filter and projections fix

squire 0.4.11

  • patching projections when projecting further than original calibrate run

squire 0.4.10

  • calibrate new argument for R0_prior

squire 0.4.9

  • calibrate now allows intervetnion dates to be before epidemic start date

squire 0.4.8

  • calibrate argument reporting_fraction implemented now in likelihood

squire 0.4.7

  • calibrate can do 3D grid scans investigating Meff
  • plot.squire_scan will collapse n-dimensional scans using show argument

squire 0.4.6

  • Patch for particle_filter_data when only 1 row of data is provided

squire 0.4.5

  • Add functionality for running the deterministic model in javascript

squire 0.4.4

  • Patch for calibrate to ensure correct dt values provided

squire 0.4.3

  • calibrate can now take squire:::deterministic_model() as model type.
  • Output of above does not work with plotting or projections yet.

squire 0.4.2

  • Patch for plot.scan_results where log likelihood was being plotted for both

squire 0.4.1

  • Patch for passing in user defined population and matrices

squire 0.4.0

  • calibrate now uses particle filter. See README for examples (#12)
    • Dates are to be provided for timings of interventions before today (#44, #17)
    • Custom populations, contact matrices and bed capacity allowed (#73, #80, #28, #16)
    • Increased modularity for new models (#62, #61, #49)
    • Returns lieklihood of different baseline R0s and start dates (#14)
  • get_mixing_matrix takes iso3c (#84)
  • Number of plotting enhancement (#72, #69)
  • Parameter references vignette and website update (#13, #1, #5, #78)

squire 0.3.0

  • projection function for doing scenario projections

squire 0.2.3

  • format_output and calibrate orderly plotting functioanlity in

squire 0.2.2

  • calibrate now allows different seeding events (#50) and undereporting (#11)
  • New plotting functions for plot.squire_calibration
  • iso3c country codes in (#52)

squire 0.2.1

  • extract_all_outputs and extract_specific_output (#37)

squire 0.2.0

  • New minor version for iterative PR development cycle from now

squire 0.1.0

  • initial version of the model