Force constants assign file

All the optimized parameters at the end of a JOYCE run are also written in a text file named assign.dat. As for the suggested dependencies file, also the content of assign.dat can also be copied and pasted into a new JOYCE input file. Note that If this is done without any previous editing, all the force constants will be constrained to their previously optimized values, and no fitting will be performed in the second run. On the contrary, if the constraints on some IC are removed from the assign.dat file before pasting it into a new input, the new fitting will be performed only on those IC removed from the list, keeping all the force constants found in the $assign environment fixed to the values optimized in the previous fitting.

$assign
  1 =   2254.5152345     C1-C2
  2 =   2130.5117241     C2-C2
  3 =   2254.5152345     C1-C2
  4 =   3100.0797022     C1-H1
            [...] 37 =    313.3362781     H1-C1-H1
 38 =     -1.9686612     C1-C2-C2-C1_n=0
 39 =      4.3993865     C1-C2-C2-C1_n=1
 40 =      1.7642241     C1-C2-C2-C1_n=2
 41 =      7.5582350     C1-C2-C2-C1_n=3
 42 =      0.1556409     C1-C2-C2-C1_n=4
 43 =      2.1534692     H1-C1-C2-C2_n=3
 44 =      2.1534692     H1-C1-C2-C2_n=3
 45 =      2.1534692     H1-C1-C2-C2_n=3
 46 =      2.1534692     H1-C1-C2-C2_n=3
 47 =      2.1534692     H1-C1-C2-C2_n=3
 48 =      2.1534692     H1-C1-C2-C2_n=3
 $end
 
Although this procedure may seem a bit cumbersome, it can be very useful in the case of a well-different set of coordinates, such as stiff and soft IC. In this case it may be useful to fit in a first run all IC, only on the base of the optimized geometry and Hessian. Since the information used is insufficient to accurately characterized soft coordinates, a second run is necessary. In this second run, QM scans on soft IC are also read by JOYCE. Thus, to simplify the fitting procedure, the second run (i.e. the one with the QM scan data) can be performed keeping all stiff constants constrained to the values optimized in the first run. This is done by pasting into the second input file the assign.dat produce by JOYCE  in the first run, after removing all soft IC from the list therein. Further details can be found by copnsulting the tutorials availabel at the JOYCE website.[13] It is worth noticing that, despite JOYCE  adds at the end of the assigned list also the non-bonded intramolecular constants ( $\epsilon_{ij}$, $\sigma{ij}$ and the product $f_{ij}*q_i*qj$), the LJ parameters and the atomic charges are not optimized by the default settings of the program.  These should be considered as assigned parameters, even if they do not appear in the assigned list. The $\epsilon_{ij}$ parameters can be varied in the fitting procedure if the fitLJ key is activated in the JOYCEinput. In this case, those $\epsilon_{ij}$ that the user wants to keep constrained, should be explicitly listed in the assign environment.