Institut für Astronomie und AstrophysikAbteilung AstronomieSand 1, D-72076 Tübingen, GermanyNew Address! -- Neue Adresse! |
TSUM
Trapezoidal summation of the area under a curve.
Adapted from the procedure INTEG in the IUE procedure library.
Result = TSUM(y) or Result = TSUM( x, y, [ imin, imax ] )
x = array containing monotonic independent variable. If omitted, then x is assumed to contain the index of the y variable. x = lindgen( N_elements(y) ). y = array containing dependent variable y = f(x)
imin = scalar index of x array at which to begin the integration If omitted, then summation starts at x[0]. imax = scalar index of x value at which to end the integration If omitted then the integration ends at x[npts-1].
result = area under the curve y=f(x) between x[imin] and x[imax].
IDL> x = [0.0,0.1,0.14,0.3] IDL> y = sin(x) IDL> print,tsum(x,y) ===> 0.0445843 In this example, the exact curve can be computed analytically as 1.0 - cos(0.3) = 0.0446635
The area is determined of individual trapezoids defined by x[i], x[i+1], y[i] and y[i+1]. If the data is known to be at all smooth, then a more accurate integration can be found by interpolation prior to the trapezoidal sums, for example, by the standard IDL User Library int_tabulated.pro.
Written, W.B. Landsman, STI Corp. May 1986 Modified so X is not altered in a one parameter call Jan 1990 Converted to IDL V5.0 W. Landsman September 1997 Allow non-integer values of imin and imax W. Landsman April 2001 Fix problem if only 1 parameter supplied W. Landsman June 2002
[Home Page] [Software, Documentation] [IDL Documentation] [Quick Reference] [Feedback]