Re: Coding for simulation

Dear Matthew, For simulation of WT with known mean and variance I use this code (different mean and variance for males and females respectively): IF(ICALL == 4) THEN ;---- WT simulation --------------------------- IF(NEWIND /= 2) THEN DO WHILE (ABS(ETA(1)).GT.0.52) CALL SIMETA(ETA) ENDDO DO WHILE (ABS(ETA(2)).GT.0.60) CALL SIMETA(ETA) ENDDO ENDIF IF(MALE == 1) WT = THETA(1)*EXP(ETA(1)) ;WT simulation for males IF(MALE == 0) WT = THETA(2)*EXP(ETA(2)) ;WT simulation for females ENDIF $THETA 85 FIX ; 1 WT MALE $THETA 75 FIX ; 2 WT FEMALE $OMEGA 0.03 FIX ; 1 OM WT MALE $OMEGA 0.04 FIX ; 2 OM WT FEMALE $SIM (12345 NEW) The "DO WHILE"-loop is used to force the simulated ETAs to not be more than 3 standard deviations away from the typical value in the population (to avoid extreme WT values) and the "NEW" option in $SIM is required to simulate a new ETA in case the one just simulated doesn't fulfill the requirements in the "DO WHILE"-loop (is more than 3 standard deviations away from THETA). For simulation of different genotypes with known proportions, this code should work: IF(ICALL == 4)THEN ;---- Genotype simulation --------------------- IF(NEWIND /= 2) THEN CALL RANDOM (1,R) RAND = R ENDIF GENOT = 1 ; Genotype A (10%) IF(RAND <= 0.9) GENOT = 2 ; Genotype B (30%) IF(RAND <= 0.6) GENOT = 3 ; Genotype C (60%) ENDIF $SIM (12345 UNI) "CALL RANDOM" draws a random value using (in this case) the first seed number specified in $SIM and the "UNI" option in $SIM specifies that the random value should be drawn from a uniform distribution [0, 1]. Since the random value is between 0 and 1, this can be used to simulates proportions. Good luck! Åsa Johansson *~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~* Åsa M. Johansson, MSc, PhD student *~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~* Pharmacometrics Research Group Department of Pharmaceutical Biosciences Uppsala University P.O. Box 591 SE - 751 24 Uppsala SWEDEN Office: +46 18-471 43 29 Fax: +46 18-471 40 03 *~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*