The rt-bops regtest was calculating the values of this type of CV inc…

…orrectly because the actions have to be a particular order in the input file in that old version. The new reference values that I have put here are consistent with the values output by master when the actions are in the correct order. Furthermore, because of the changes to the mechanism via which the forces are applied the order of operations is no longer important

regtest/crystdistrib/rt-bops/colvar.reference

@@ -1,4 +1,4 @@
 #! FIELDS time s
- 0.000000    10.66
- 0.025000   922.28
- 0.050000   993.43
+ 0.000000   925.50
+ 0.025000  1019.81
+ 0.050000  1064.21

regtest/crystdistrib/rt-bops/config

@@ -2,3 +2,11 @@ type=driver
 plumed_modules="adjmat crystdistrib"
 arg="--plumed plumed.dat --mf_xtc em.xtc --mc mcfile --timestep=0.001 --trajectory-stride=25"
 extra_files="../../trajectories/em.xtc" 
+
+function plumed_regtest_before() {
+   export PLUMED_FORBID_CHAINS=yes
+}
+
+function plumed_regtest_after() {
+   export PLUMED_FORBID_CHAINS=no
+}

regtest/crystdistrib/rt-bops/plumed.dat

@@ -260,6 +260,8 @@ ops: CUSTOM ARG=qp.i,qp.j,qp.k,c1.w VAR=i,j,k,d FUNC=((2.26628554234*exp(-(32.27
 
 ones: ONES SIZE=252
 
+cc: MATRIX_VECTOR_PRODUCT ARG=qp.i,ones
+PRINT ARG=cc FILE=cc FMT=%8.4f 
 
 sdot: MATRIX_VECTOR_PRODUCT ARG=ops,ones
 s: SUM ARG=sdot PERIODIC=NO