Initial commit

2016-02-03 18:52:05 +00:00
commit d40505e161
507 changed files with 91383 additions and 0 deletions
@@ -0,0 +1,83 @@
+      subroutine histogram(npoints,datalabel,datavalue,ngroup,
+     &grouplabel,nhistmark,histmark,nfreq,ncumumark,cumumark,
+     &ncumu)
+      implicit none
+!npoints: the total number of points of the whole series
+!datalabel: the lable of each datum
+!datavalue: the value of each datum
+!ngroup: the number of different data labeles, each label represents a group
+!grouplabel: the label of each group
+!nhistmark: the number of histogram band marks in each group
+!histmark: the value of each histogram mark
+!nfreq: the number of points in each histogram bands
+!ncumumark: the number of marks for cumulative distribution
+!cumumark: the marks for the cumulative distribution
+!ncumu: the cumulative distribution in each group
+      integer npoints,ngroup,nfreq(ngroup,npoints),nhistmark(ngroup),
+     &ncumumark(ngroup),ncumu(ngroup,npoints),iorder(npoints),i,j,n
+      character*50 datalabel(npoints),grouplabel(ngroup)
+      double precision datavalue(npoints),histmark(ngroup,npoints),
+     &cumumark(ngroup,npoints),vector(npoints)
+      do i=1,ngroup
+        do j=1,npoints
+          nfreq(i,j)=0
+          ncumu(i,j)=0
+        enddo
+        ncumumark(i)=0
+      enddo
+      do k=1,npoints
+        do i=1,ngroup
+          if(trim(datalabel(k)).eq.trim(grouplabel(i))then
+!initially we set the ncumumark to the total number of points in each group. Later we will
+!merge points with the same value
+            ncumumark(i)=ncumumark(i)+1
+            cumumark(i,ncumumark(i))=datavalue(k)
+            goto 10
+          endif
+        enddo
+10      continue
+      enddo
+      do i=1,ngroup
+        call sort_shell(ncumumark(i),cumumark(i:i,1:ncumumark(i)),
+     &iorder)
+        do k=1,ncumumark(i)
+          do j=1,nhistmark(i)-1
+            if(j.eq.(nhistmark(i)-1))then
+              if(cumumark(i,k).ge.histmark(i,j).and.cumumark(i,k).le.
+     &histmark(i,j+1))then
+              nfreq(i,j)=nfreq(i,j)+1
+              goto 20
+            else
+              if(cumumark(i,k).ge.histmark(i,j).and.cumumark(i,k).lt.
+     &histmark(i,j+1))then
+                nfreq(i,j)=nfreq(i,j)+1
+                goto 20
+              endif
+            endif
+          enddo
+20        continue
+        enddo
+!cumulative distribution. we have to merge points with equal values
+        do j=1,ncumumark(i)
+          vector(j)=cumumark(i,j)
+        enddo
+        k=ncumumark(i)
+        ncumumark(i)=1
+        cumumark(i,1)=vector(1)
+        do j=2,k
+          if(vector(j).ne.cumumark(i,ncumumark(i)))then
+            ncumumark(i)=ncumumark(i)+1
+            cumumark(i,ncumumark(i))=vector(j)
+          endif
+        enddo
+        do j=1,ncumumark(i)
+          ncumu(i,j)=0
+          do n=1,k
+            if(ivector(n).le.cumumark(i,j))then
+              ncumu(i,j)=ncumu(i,j)+1
+            endif
+          enddo
+        enddo
+      enddo
+      return
+      end