Welcome. This is CARS.jl. If you haven't read the publication yet, please do so. It will answer many of the questions that you may have about what this code does and why: https://doi.org/10.1002/jrs.6639
CARS.jl is a library-based fitting algorithm for coherent anti-Stokes Raman spectra that features:
- Dual-Pump/Single-Pump evaluation
- High processing speed
- High accuracy
- Automatic determination of wavenumber axis using a 2nd order polynomial fitted to the actual data
- In the libraries, the only tabulated parameter is temperature, thus the size grows (nearly) linearly - not with the power of - the number of species
- generate final spectra from theoretical susceptibilities supplied by other packages, such as DIACars.jl
- generate libraries of complex susceptibilities using a loss-less compression scheme for fitting
- fit spectra to experimental data
- supply several noise models based on shot-noise, Stokes noise and possibly custom noise-models for weighting the residual in the non-linear least squares fit
- store evaluated data in either JLD2 or MAT format for further data evaluation
- Generate theoretical susceptibilities. This has to be done using other packages such as DIACARS.jl. This modular structure allows to easily include additional species.
- CARS.jl assumes independency of theoretical susceptibilities of the mole fractions of other species in the probe volume. This implies that foreign gas broadening is neglected. This is a common approximation and is e.g. also used by CARSFT
- At present, only fitting at constant pressure are possible. Including pressure in the fit requires introducing pressure variations in the library, because it affects the underlying line shapes.
- Single Pump CARS is not directly supported at the moment. Simply pass the same library twice (as Single-Pump is actually a special case of Dual-Pump) to the fit/simulation to workaround this limitation. This is likely to be resolved in a future version but does not have the highest priority.
