DocumentCode :
333089
Title :
LABATCH.2: software for statistical analysis of simulation sample path data
Author :
Fishman, George S.
Author_Institution :
Dept. of Oper. Res., North Carolina Univ., Chapel Hill, NC, USA
Volume :
1
fYear :
1998
fDate :
13-16 Dec 1998
Firstpage :
131
Abstract :
LABATCH.2 is a collection of computer programs available in C, FORTRAN, and SIMSCRIPT II.5. It performs statistical analyses on sample sequences collected on strictly stationary stochastic processes. Designed to make its implementation easy for potential users, it may be invoked in-line or from a stored data file. For each sample sequence of length t, LABATCH.2 takes O(t) computing time and O(log2t) space. For each series, LABATCH.2 provides an asymptotically valid confidence interval, based on the batch means method, for assessing how well its sample average approximates its true unknown mean. It also produces interim estimates of the variance of the sample average that enable a user to detect systematic error in the latest variance estimate, due to correlation between batches. It also allows a user to assess the extent to which the sample average is free of initial conditions. LABATCH.2 has an interactive option that displays interim results on screen. Based on these quantities, a user may instruct LABATCH.2 to continue execution until the next update or to terminate statistical analysis and write the final results to a file
Keywords :
digital simulation; software packages; statistical analysis; LABATCH.2; batch means method; computing time; interactive option; simulation sample path data; stationary stochastic processes; statistical analysis; Algorithms; Analytical models; Buildings; Computational modeling; Data analysis; Displays; Operations research; Process design; Statistical analysis; Stochastic processes;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Simulation Conference Proceedings, 1998. Winter
Conference_Location :
Washington, DC
Print_ISBN :
0-7803-5133-9
Type :
conf
DOI :
10.1109/WSC.1998.744908
Filename :
744908
Link To Document :
بازگشت