From: oetiker <oetiker@a5681a0c-68f1-0310-ab6d-d61299d08faa>
Date: Thu, 4 Aug 2005 05:08:04 +0000 (+0000)
Subject: New VDEF functions for least squares slope, intercept & correlation for simple foreca... 
X-Git-Url: https://git.tokkee.org/?a=commitdiff_plain;h=1691631da4508df9e9d97cd720f7cf95065f94f6;p=rrdtool-all.git

New VDEF functions for least squares slope, intercept & correlation for simple forecasting
-- Trent Burkard <tburkard with tangentis.com>


git-svn-id: svn://svn.oetiker.ch/rrdtool/branches/1.2@669 a5681a0c-68f1-0310-ab6d-d61299d08faa
---

diff --git a/program/doc/rrdgraph_rpn.pod b/program/doc/rrdgraph_rpn.pod
index 74ac4ad4..1aaef653 100644
--- a/program/doc/rrdgraph_rpn.pod
+++ b/program/doc/rrdgraph_rpn.pod
@@ -272,6 +272,18 @@ finite numbers and are always more than the I<Unknown> numbers.
 
 Example: C<VDEF:perc95=mydata,95,PERCENT>
 
+=item LSLSLOPE, LSLINT, LSLCORREL
+
+Return the parameters for a B<L>east B<S>quares B<L>ine I<(y = mx +b)> 
+which approximate the provided dataset.  LSLSLOPE is the slope I<(m)> of
+the line related to the COUNT position of the data.  LSLINT is the 
+y-intercept I<(b)>, which happens also to be the first data point on the 
+graph. LSLCORREL is the Correlation Coefficient (also know as Pearson's 
+Product Moment Correlation Coefficient).  It will range from 0 to +/-1 
+and represents the quality of fit for the approximation.   
+
+Example: C<VDEF:slope=mydata,LSLSLOPE>
+
 =back
 
 =head1 SEE ALSO
diff --git a/program/src/rrd_graph.c b/program/src/rrd_graph.c
index 2a13cb54..3ce39598 100644
--- a/program/src/rrd_graph.c
+++ b/program/src/rrd_graph.c
@@ -3406,6 +3406,9 @@ char *str;
     else if	(!strcmp("TOTAL",  func)) gdes->vf.op = VDEF_TOTAL;
     else if	(!strcmp("FIRST",  func)) gdes->vf.op = VDEF_FIRST;
     else if	(!strcmp("LAST",   func)) gdes->vf.op = VDEF_LAST;
+    else if     (!strcmp("LSLSLOPE", func)) gdes->vf.op = VDEF_LSLSLOPE;
+    else if     (!strcmp("LSLINT",   func)) gdes->vf.op = VDEF_LSLINT;
+    else if     (!strcmp("LSLCORREL",func)) gdes->vf.op = VDEF_LSLCORREL;
     else {
 	rrd_set_error("Unknown function '%s' in VDEF '%s'\n"
 	    ,func
@@ -3441,6 +3444,9 @@ char *str;
 	case VDEF_TOTAL:
 	case VDEF_FIRST:
 	case VDEF_LAST:
+	case VDEF_LSLSLOPE:
+	case VDEF_LSLINT:
+	case VDEF_LSLCORREL:
 	    if (isnan(param)) {
 		gdes->vf.param = DNAN;
 		gdes->vf.val   = DNAN;
@@ -3599,6 +3605,48 @@ printf("DEBUG: %3li:%10.2f %c\n",step,array[step],step==field?'*':' ');
 		dst->vf.when = src->start + (step+1)*src->step;
 	    }
 	    break;
+	case VDEF_LSLSLOPE:
+	case VDEF_LSLINT:
+	case VDEF_LSLCORREL:{
+	    /* Bestfit line by linear least squares method */ 
+
+	    int cnt=0;
+	    double SUMx, SUMy, SUMxy, SUMxx, SUMyy, slope, y_intercept, correl ;
+	    SUMx = 0; SUMy = 0; SUMxy = 0; SUMxx = 0; SUMyy = 0;
+
+	    for (step=0;step<steps;step++) {
+		if (finite(data[step*src->ds_cnt])) {
+		    cnt++;
+		    SUMx  += step;
+		    SUMxx += step * step;
+		    SUMxy += step * data[step*src->ds_cnt];
+		    SUMy  += data[step*src->ds_cnt];
+		    SUMyy  += data[step*src->ds_cnt]*data[step*src->ds_cnt];
+		};
+	    }
+
+	    slope = ( SUMx*SUMy - cnt*SUMxy ) / ( SUMx*SUMx - cnt*SUMxx );
+	    y_intercept = ( SUMy - slope*SUMx ) / cnt;
+	    correl = (SUMxy - (SUMx*SUMy)/cnt) / sqrt((SUMxx - (SUMx*SUMx)/cnt)*(SUMyy - (SUMy*SUMy)/cnt));
+
+	    if (cnt) {
+		    if (dst->vf.op == VDEF_LSLSLOPE) {
+			dst->vf.val  = slope;
+			dst->vf.when = cnt*src->step;
+		    } else if (dst->vf.op == VDEF_LSLINT)  {
+			dst->vf.val = y_intercept;
+			dst->vf.when = cnt*src->step;
+		    } else if (dst->vf.op == VDEF_LSLCORREL)  {
+			dst->vf.val = correl;
+			dst->vf.when = cnt*src->step;
+		    };
+		
+	    } else {
+		dst->vf.val  = DNAN;
+		dst->vf.when = 0;
+	    }
+	    }
+	    break;
     }
     return 0;
 }
diff --git a/program/src/rrd_graph.h b/program/src/rrd_graph.h
index 9ced52ee..01b6799c 100644
--- a/program/src/rrd_graph.h
+++ b/program/src/rrd_graph.h
@@ -42,6 +42,9 @@ enum vdef_op_en {
 		,VDEF_TOTAL	/* average multiplied by time */
 		,VDEF_FIRST	/* first non-unknown value and time */
 		,VDEF_LAST	/* last  non-unknown value and time */
+		,VDEF_LSLSLOPE  /* least squares line slope */
+		,VDEF_LSLINT    /* least squares line y_intercept */
+		,VDEF_LSLCORREL /* least squares line correlation coefficient */
 		};
 enum text_prop_en { TEXT_PROP_DEFAULT=0,   /* default settings */
 	            TEXT_PROP_TITLE,       /* properties for the title */