Fixed these two regtests

These two tests needed changing because periodicity was not incorporated correctly
in the KDE estimation previously.  As such the outputs given were wrong previously.
The previous changed incorporated the periodicity correctly.  This change is thus
required as a result.

regtest/analysis/rt-uweights-integral/colvar.reference

@@ -3,4 +3,4 @@
  1.000000   0.5437   0.2128   4.2565
  2.000000   0.0000   2.8124  56.2470
  3.000000   0.0000   2.8122  56.2440
- 4.000000   2.0008   7.8226 156.4529
+ 4.000000   2.0008   7.8230 156.4606

regtest/analysis/rt-uweights-integral/forces.reference

@@ -811,9 +811,9 @@ X   0.0000   0.0000   0.0000
 X   0.0000   0.0000   0.0000
 X   0.0000   0.0000   0.0000
 201
- -0.0689   0.0557   0.0132
-X   0.0711   0.0893  -0.0108
-X  -0.0711  -0.0893   0.0108
+ -0.0650   0.0517   0.0132
+X   0.0676   0.0839  -0.0108
+X  -0.0676  -0.0839   0.0108
 X   0.0266   0.0072  -0.0013
 X  -0.0000  -0.0000   0.0000
 X  -0.0266  -0.0072   0.0013

regtest/analysis/rt-weights-integral/colvar.reference

@@ -3,4 +3,4 @@
  1.000000   0.9993   0.3108   6.2163
  2.000000   1.0002   0.3130   6.2595
  3.000000   0.9888   0.2852   5.7033
- 4.000000   1.0004   0.3135   6.2703
+ 4.000000   1.0004   0.3136   6.2711

regtest/analysis/rt-weights-integral/forces.reference

@@ -811,9 +811,9 @@ X   0.0000   0.0000   0.0000
 X   0.0000   0.0000   0.0000
 X   0.0000   0.0000   0.0000
 201
- -0.0069   0.0056   0.0013
-X   0.0071   0.0089  -0.0011
-X  -0.0071  -0.0089   0.0011
+ -0.0065   0.0052   0.0013
+X   0.0068   0.0084  -0.0011
+X  -0.0068  -0.0084   0.0011
 X   0.0027   0.0007  -0.0001
 X   0.0000   0.0000  -0.0000
 X  -0.0027  -0.0007   0.0001