Page 1 of 1

Too many tokens supplied. Please increase the mdet_max_mods

Posted: Mon Jul 31, 2017 11:51 am
by s.nasiri
Dear All
I generate a WFN for anionic form of Lithium with MCSCF(4,9) mathod. the number of determinants are about 400. I got the following error:
#####################################
ERROR : READ_TOKENS
Too many tokens supplied. Please increase the mdet_max_mods parameter in
mdet.f90.

CASINO internal traceback:
Problem detected at SETUP
Called from MAIN
#####################################
there is no mdet.f90 file, but the ABP.f90 file has the message (src directory). I increased the mdet_max_mods (=5000) and compile the CASINO , but it does not work again (with the same error).
Is it rigth to incrase the mdet_max_mods from 1000 (defult) to larger value (5000 or 20000)?
What should I do to solve this error?

Thanks.

Re: Too many tokens supplied. Please increase the mdet_max_m

Posted: Mon Jul 31, 2017 12:47 pm
by Pablo_Lopez_Rios
Hi s.nasiri,

Since v2.13.617 (4-July-2017) mdet_max_mods has been increased to 10,000 in the CASINO source. This value gives the length of a fixed-size vector, which you can increase reasonably without side effects -- the memory used by the vector in question is linear in this number, so you are unlikely to run into memory trouble for changing its size. I would suggest using 10,000 -- 100,000, which should be plenty if your 400-determinant expansion is dominated by double excitations from the reference.

When I find time I will rewrite the relevant bit of code so that we use dynamic allocation in the MDET parser.

Best,
Pablo

Re: Too many tokens supplied. Please increase the mdet_max_m

Posted: Fri Oct 06, 2017 7:42 am
by Pablo_Lopez_Rios
Hi,
Pablo_Lopez_Rios wrote:When I find time I will rewrite the relevant bit of code so that we use dynamic allocation in the MDET parser.
This rewrite is now in CASINO as of v2.13.650; this error should not appear again.

Best,
Pablo