v0.4.2

Updates

• examples

  • Renamed core functions

• docs

  • Fix example codes

• Parameter estimation

  • Add new local search method: "CMAES"

    • The CMA Evolution Strategy is a stochastic method for derivative-free optimization of potentially non-linear, non-convex or noisy functions over continuous domains (Hansen 2016).

  • Fix search bounds in SciPyOptimize

v0.4.1

Update visualize

• save_format::String (default: "pdf")
  • Either "png" or "pdf", indicating whether to save figures as png or pdf format.

Update optimize

• initial_threshold::Float64 (default: 1e12)
  • Allowable error used to generate initial population. Default is 1e12 (numerically solvable).

v0.4.0 (January 2021)

• Add a Delay differential Equation (DDE) model

using BioMASS

dde_model = load_model("examples/nfkb_model");

visualize(dde_model, viz_type="original")

v0.3.4

• Add new function to visualize estimated parameter region
• Allow users to modify the maximum number of iterations for the local search methods: "Powell" and "DE"

v0.3.3

• Update docs
• Remove Random from deps

v0.3.2

Updates:

• Add new local search method: "DE"

  • using differential evolution with 'best2bin' strategy in local search phase

Example usage

using BioMASS

model = load_model("./examples/fos_model")

optimize(model, 1, max_generation=1000, allowable_error=0.35, local_search_method="DE")

v0.3.1

Updates:

• Add new local search method: "powell"
  • Using modified Powell's method (scipy.optimize.minimize(method="Powell"))

v0.3.0 (December 2020)

Updates:

• Documentation

New Features:

• Bifurcation analysis

v0.2.1

Fix index in set_search_param.jl

v0.2.0 (November 2020)

Updates:

• Create visualize.jl

  visualize(model, viz_type="best", show_all=true)

  • viz_type : String

    • "average":
      The average of simulation results with parameter sets in "fitparam/".
    • "best":
      The best simulation result in "fitparam/".
    • "original":
      Simulation with the default parameters and initial values defined in set_model.jl.
    • "n(=1,2,...)":
      Use the parameter set in "fitparam/n/".

  • show_all : Bool

    • Whether to show all simulation results.

  • stdev : Bool

    • If True, the standard deviation of simulated values will be shown (only available for "average" visualization type).

• Change Sim.normalization type from Bool to Dict:

  normalization = Dict{String, Dict{}}()
  for observable in observables
      normalization[observable] = Dict(
          "timepoint" => nothing,
          "condition" => []
      )
  end

  • "timepoint" : Union{Int, Nothing}

    • The time point at which simulated values are normalized. If nothing, the maximum value will be used for normalization.

  • "condition" : Vector{Stirng}

    • The experimental conditions to use for normalization. If empty, all conditions defined int Sim.conditions will be used.