Version history
IDAP 1.0
- Functions from Global Optimization toolbox are included as minimization algorithms. Their options may be set through corresponding properties of the TotalFit object
- Utilization of parallel processors for Global Optimization functions and Monte-Carlo (when run on a multi-core workstation) is included (Parallel Processing Toolbox is required). Distributed Monte-Carlo's using deamons still utilize deamons running on individual cores (due to MathWorks limitations).
- Modes were renamed for the fit(mode) function to be 'single_cluster' and 'multiple_cores'.
- Synthax of model functions calls (parameter lists) is changed: include now the name of a solver and its options. Models that are solved numerically will use these parameters. Others will ignore them.
To migrate your own existing models you created for previous version of IDAP---simply include to extra parameters in the model function call (and not use them if you don't want):
-
In code/+equilibrium_thermodynamic_equations/:
U_model(Rtotal, LRratio, K_A, model_numeric_solver, model_numeric_options);
- In code/+line_shape_equations_1D/:
U_model_1D(parameters, X, model_numeric_solver, model_numeric_options)
- Transient kinetics module RapidMixingKinetics is included.
- The code for creating kinetic matrices for NMR line shapes was moved into code/+line_shape_equations_1D. The folder +line_shape_kinetic_matrices was removed.
- The following methods now require explicit solver and solver options given along with the model:
- Dataset: set_active_model()
- TotalFit: dataset_set_active_model()
- TotalFit: change_model_in_series()
- TotalFit: create_simulation_series()
- output functions output_figure() and re_export_results() are moved to code/+results_output
- Output arguments of functions calculating equlibrium concentration of species changed in +equilibrium_thermodynamic_equations/. Now it returns a single vector of concentrations and a single cell array of names of species.
- The 1D line shape models no longer normalize concentrations by total concentration of the receptor. Instead, ScaleFactor will take care of normalization in the model. I did it because some models will include two different receptors (a model with mixture)---then this normalization becomes ambiguous. As result, spectral intensity now takes values in a range of 1e-4.