speed up wrf_create

[atmoschem]

speed up wrf_create

[Schuch666]

wrf_create is not a particularly slow function in general, but can take a minute or more depending on the number of times (frames_per_auxinput5) and number of emission species

There is two ways to speed up this process:

• reduce the number of times on the file
• reduce the number of variables (WRF ignores the missing variables), I recommend to create the emission without the unused variables using a subset of the emiss_opt variables of the WRF option

[ibarraespinosa]

Yeah, but line 99 has de for. So maybe, using a applay* could speed up this process. I will make some tests.

Another approach is parallelize code with doMC

This means that we have 4 approaches.

[Schuch666]

Yes indeed, but reduce the number of times or reduce the number of variables need just the change of 1 or 2 arguments on the function call

In the line 99 the loop are for multiple files, maybe doMC can make some work here (but its add more one dependence to the package)

A god place to start the 'apply type' functions are the following loops, that declare the NECTDF variables
GAS phase emissions
AEROSOL emissions

I recommend caution with a possible 'parallel file writing' on the following loops
GASES initialization with zeros
AEROSOLS initialization with zeros

PS: You don't need to create new files for different scenarios (same period, domain and chem options), can just check if the emission already exists