Log in
Register|
d_labes ★★★ Berlin, Germany, 2010-04-19 15:41 (5900 d 05:15 ago) Posting: # 5168 Views: 12,063 |
|
|
Dear All, just discovered in Frank Harrels library (package) Hmisc: Area under curve defined by arbitrary x and y vectors, using trapezoidal rule and assuming x sorted appropriately trap.rule <- function(x, y) sum(diff(x) * (y[-1] + y[-length(y)]))/2Wow  ! The cream of the crop of programming art!— Regards, Detlew |
|
Helmut ★★★   Vienna, Austria, 2010-04-20 14:53 (5899 d 06:03 ago) @ d_labes Posting: # 5173 Views: 10,233 |
|
|
Dear D. Labes, another option (besides bear, of course) in package PK by Thomas Jaki and Martin J. Wolfsegger: time <- c(0,0.25,0.5,0.75,1,1.5,2,2.5,3,4,5,6)gives Estimation for a complete data design— Dif-tor heh smusma 🖖🏼 Довге життя Україна! ![[image]](https://static.bebac.at/pics/Blue_and_yellow_ribbon_UA.png) Helmut Schütz ![[image]](https://static.bebac.at/img/CC by.png) The quality of responses received is directly proportional to the quality of the question asked. 🚮 Science Quotes |
|
d_labes ★★★ Berlin, Germany, 2010-04-20 16:35 (5899 d 04:21 ago) @ Helmut Posting: # 5174 Views: 10,076 |
|
|
Dear Helmut, ❝ ❝ ❝ [code] AUC AUMC ❝ ❝ ❝ How did you do that  .If I run your code I always get only an error: > auc(conc,time,design='complete')Used by me: PK Version 1.2-1 R 2.10.0 If I specify a method in the auc() call I get > auc(conc, time, group=NULL, design='complete',method="t")— Regards, Detlew |
|
Jack ☆ Lancaster, United Kingdom, 2010-04-20 17:21 (5899 d 03:34 ago) @ d_labes Posting: # 5175 Views: 10,130 |
|
|
Dear D. Labes, as the maintainer of the PK package I guess I need to say something here. The output Helmut got was created with Version 1.01 of our package. Since then substantial improvements to the package have been made which (among other things) has led to a different output format being used. In addition Versions of the package > 1.01 by default also compute the standard error of the AUC when using the auc function. As the example only contains data of one subject this variability is zero. Using the function call without specifying the method means that a bootstrap interval will be computed which, due to the lack of variability, results in the error. Specifying the methods "t" or "z", however, gives the point estimate but a SE of zero and no confidence interval. Best, Jack |
|
Helmut ★★★ Vienna, Austria, 2010-04-20 17:31 (5899 d 03:24 ago) @ d_labes Posting: # 5176 Views: 10,016 |
|
|
Dear D. Labes, oops, sorry. Used the previous version on R2.9.0. Could reproduce the message with PK1.2 on R2.10.1. Since I'm riding at 200km/h through one tunnel after the other I'm loosing my internet connection all the time. See help(nca) for the new syntax.— Dif-tor heh smusma 🖖🏼 Довге життя Україна! Helmut Schütz The quality of responses received is directly proportional to the quality of the question asked. 🚮 Science Quotes |
|
Jack ☆ Lancaster, United Kingdom, 2010-04-23 19:00 (5896 d 01:56 ago) @ d_labes Posting: # 5204 Views: 10,092 |
|
|
Dear D. Labes, We have just released a new version of the PK package (1.2-2) which, among other things, now allows a more natural estimation of individual PK parameters. Besides time <- c(0,0.25,0.5,0.75,1,1.5,2,2.5,3,4,5,6)which produces a point estimate of the AUC one now can also use for example nca(conc,time, n.tail=3, dose=100, design='complete')to obtain other parameters of interest. Note that both functions do not provide standard errors and confidence intervals at the moment. Best, Jack |
|
d_labes ★★★ Berlin, Germany, 2010-04-26 10:04 (5893 d 10:51 ago) @ Jack Posting: # 5225 Views: 9,967 |
|
|
Dear Jack, ❝ We have just released a new version of the PK package (1.2-2) Thanks for your announcement. — Regards, Detlew |
Ing. Helmut Schütz