Skip to content

Commit

Permalink
'labeling_related' improved a number of comments. One was breaking fo…
Browse files Browse the repository at this point in the history
…rtpy because it was using the <> symbols
  • Loading branch information
Gus Hart committed Sep 10, 2019
1 parent 9dfc347 commit 20d7bae
Showing 1 changed file with 6 additions and 6 deletions.
12 changes: 6 additions & 6 deletions src/labeling_related.f90
Original file line number Diff line number Diff line change
Expand Up @@ -514,7 +514,7 @@ END SUBROUTINE generate_permutation_labelings
!!<summary>This routine takes in a list of labels of the parent cell
!!and the number of different labels allowed on each site. It
!!returns a "multiplier", and *expanded* versions of parLabel and
! parDigit.</summary>
!! parDigit.</summary>
!!<parameter name="n" regular="true">The index (volume factor) of
!!the supercell.</parameter>
!!<parameter name="k" regular="true">k-nary case</parameter>
Expand Down Expand Up @@ -1186,7 +1186,7 @@ FUNCTION labeling_is_legal(labeling,siteLabels,digitN)
!!super-periodic labelings (non-primitive superstructures). If the
!!"full" variable is false, it also removes "label-permutation"
!!duplicates---labelings that are not unique when the labels
!!themselves (not their positions) are permuted (e.g., 00111 &lt;--&gt;
!!themselves (not their positions) are permuted (e.g., 00111 turns into
!!11000). The basic idea of the routine is to run like an
!!"odometer", generating all numbers (base k) from 0 to k^n - 1, and
!!then use rotation and translation permutations to eliminate
Expand Down Expand Up @@ -1301,8 +1301,8 @@ SUBROUTINE generate_unique_labelings(k,n,nD,perm,full,lab,parLabel,parDigit,dege
!call make_translation_group(d,trgrp) ! Find equivalent
! translations (permutations of labelings)

ic = 0; c = 0; c(0) = nl ! Loop counter for fail safe; initialize
! digit counter In the recent version of multienum.x, we can't assume
ic = 0; c = 0; c(0) = nl ! Loop counter for fail safe; initialize digit counter
! In the recent version of multienum.x, we can't assume
! that the label of the first type appears on all sites. So this naive
! initialization doesn't work any more. Instead we need to count how
! many times each label appears in the starting labeling. I think that
Expand All @@ -1318,7 +1318,7 @@ SUBROUTINE generate_unique_labelings(k,n,nD,perm,full,lab,parLabel,parDigit,dege

ic = 0
nUniq = 0
do; ic = ic + 1
do; ic = ic + 1 ! Main loop over odometer readings
if (ic > nexp) exit ! Fail safe
idx = sum((digCnt-1)*multiplier)+1
if (any(c==0)) then ! Check to see if there are missing digits
Expand All @@ -1342,7 +1342,7 @@ SUBROUTINE generate_unique_labelings(k,n,nD,perm,full,lab,parLabel,parDigit,dege
cycle
endif
idx = sum((digCnt(perm(q,:))-1)*multiplier)+1
if (.not. fixed_cells) then
if (.not. fixed_cells) then ! (we don't eliminate superperiodic cases in 'fixed_cells' mode)
! This will happen if the coloring is superperiodic
! (i.e., non-primitive superstructure). The q=<n
! condition makes sure we are considering a
Expand Down

0 comments on commit 20d7bae

Please sign in to comment.