Anu ★ India, 2013-11-25 07:06 (4245 d 03:57 ago) Posting: # 11956 Views: 8,326 |
|
Dear all, Greetings! Kindly check the link Stab in R Had to found the extrapolated values for a stability data which I have calculated. Please help me knowing,how they have coded in R for the Lower and upper confidence bound, graph and the shelf life for the batches. Thanks & Regards Anu Jaswal |
yjlee168 ★★★ ![]() ![]() Kaohsiung, Taiwan, 2013-11-25 07:45 (4245 d 03:17 ago) @ Anu Posting: # 11957 Views: 7,558 |
|
Dear Anu, ❝ [...] Had to found the extrapolated values for a stability data which I have calculated. What 'extrapolated values' did you calculate? don't quite understand your questions. ❝ Please help me knowing,how they have coded in R for the Lower and upper confidence bound, graph and the shelf life for the batches. If you like to access all source codes, you can check with R website for packages. It's open sourced. Or here for more details. ❝ [...] — All the best, -- Yung-jin Lee bear v2.9.2:- created by Hsin-ya Lee & Yung-jin Lee Kaohsiung, Taiwan https://www.pkpd168.com/bear Download link (updated) -> here |
Anu ★ India, 2013-11-25 08:22 (4245 d 02:40 ago) @ yjlee168 Posting: # 11958 Views: 7,446 |
|
Dear Yung-jin Lee, Thankyou so much for the reply. As in the link, The initial values for the Batches was from (0-36 months). Extrapolated the values to (0-85 months). ❝ What 'extrapolated values' did you calculate? Was not able to code in R, to find the lower and upper confidence bound, and the graph as given in the link. Thanks & Regards Anu |
yjlee168 ★★★ ![]() ![]() Kaohsiung, Taiwan, 2013-11-25 08:34 (4245 d 02:28 ago) @ Anu Posting: # 11959 Views: 7,432 |
|
Dear Anu, ❝ [...] As in the link, The initial values for the Batches was from ( 0-36 months). ❝ Extrapolated the values to( 0-85 months). I see. It's just a "predicted" shelf-life obtained from stab for R, even though you have the initial data for only 0-36 months. Finally, you will still need the "real" lab data to confirm the "predicted" shelf-life as claimed (such as shelf-life = 60 months) officially. That's what stab for R is designed for. It provides the information of "predicted" shelf-life to allow the manufacturer to decide how long the long-term stability should be conducted. ❝ ❝ What 'extrapolated values' did you calculate? ❝ ❝ Was not able to code in R,to find the lower and upper confidence bound, and the graph as given in the link. You confused me again. Yes, it is able to code in R and should be very easy to understand the whole scheme. — All the best, -- Yung-jin Lee bear v2.9.2:- created by Hsin-ya Lee & Yung-jin Lee Kaohsiung, Taiwan https://www.pkpd168.com/bear Download link (updated) -> here |
jag009 ★★★ NJ, 2013-11-25 18:34 (4244 d 16:28 ago) @ Anu Posting: # 11962 Views: 7,408 |
|
Hi, What do you mean by "not able to code in R"? The program takes care of everything and you just need to provide input (specify upper or lower or both confidence bounds). Unless you want to change from 95% to something else then you would need to alter the R codes? John |
yjlee168 ★★★ ![]() ![]() Kaohsiung, Taiwan, 2013-11-25 19:06 (4244 d 15:57 ago) @ jag009 Posting: # 11965 Views: 7,361 |
|
Hi John, Thanks for your message. It reminds me one thing. Two year ago, one user asked if I can changed the current CI from 95% to 99% in stab for R. So I built a 'special version' for him with 99% CI. Do you know that if it is common practice in dealing with data analysis of drug stability to use 95% CI or 99% CI? I cannot find it from available resources. If so, I may consider to leave that as an option later for user to change (the default is still 95%) if necessary. ❝ What do you mean by "not able to code in R"? The program takes care of everything and you just need to provide input (specify upper or lower or both confidence bounds). Unless you want to change from 95% to something else then you would need to alter the R codes? — All the best, -- Yung-jin Lee bear v2.9.2:- created by Hsin-ya Lee & Yung-jin Lee Kaohsiung, Taiwan https://www.pkpd168.com/bear Download link (updated) -> here |
jag009 ★★★ NJ, 2013-11-25 21:21 (4244 d 13:41 ago) @ yjlee168 Posting: # 11966 Views: 7,424 |
|
❝ Do you know that if it is common practice in dealing with data analysis of drug stability to use 95% CI or 99% CI? I cannot find it from available resources. If so, I may consider to leave that as an option later for user to change (the default is still 95%) if necessary. Hi Ujlee, Yes it is still 95% as far as I know for FDA but not sure about other countries. John |
yjlee168 ★★★ ![]() ![]() Kaohsiung, Taiwan, 2013-11-26 12:03 (4243 d 22:59 ago) @ jag009 Posting: # 11970 Views: 7,381 |
|
Hi John, Thanks for your messages. I have decided to leave CI (%) as an option for next release (95% CI is still the default). ❝ Yes it is still 95% as far as I know for FDA but not sure about other countries. — All the best, -- Yung-jin Lee bear v2.9.2:- created by Hsin-ya Lee & Yung-jin Lee Kaohsiung, Taiwan https://www.pkpd168.com/bear Download link (updated) -> here |
Anu ★ India, 2013-11-29 08:18 (4241 d 02:44 ago) @ jag009 Posting: # 11988 Views: 7,242 |
|
Hi Yung-jin Lee & John, ❝ What do you mean by "not able to code in R"? The program takes care of everything and you just need to provide input (specify upper or lower or both confidence bounds). Unless you want to change from 95% to something else then you would need to alter the R codes? Sorry for the confusion, I did it for the firt time. Package itself is doing everything :) I thought I have to do all the coding by myself. Thanks & Regards Anu Jaswal |
yjlee168 ★★★ ![]() ![]() Kaohsiung, Taiwan, 2013-11-29 08:29 (4241 d 02:33 ago) @ Anu Posting: # 11989 Views: 7,225 |
|
Dear Anu OK, good to hear that. No, you don't need to do any coding. It's all menu-driven. I have uploaded stab for R v.0.1.7 to CRAN last night. And it will be available for update within 24 hr. Now user can change % CI if required (the default is still 95% CI). I know you don't need to change this option. Still you can upgrade to v0.1.7 with the command "update.packages()" under R console. ❝ Sorry for the confusion, I did it for the firt time. Package itself is doing everything :) I thought I have to do all the coding by myself... — All the best, -- Yung-jin Lee bear v2.9.2:- created by Hsin-ya Lee & Yung-jin Lee Kaohsiung, Taiwan https://www.pkpd168.com/bear Download link (updated) -> here |