udayk
☆    

India,
2020-03-06 10:00
(1781 d 19:06 ago)

Posting: # 21214
Views: 8,902
 

 Outlier Test using R-Software and Phoenix WinNonlin software in the BE [Outliers]

Post reply
Dear All,

Please help us for outlier test using Phoenix WinNonlin Software (Studentized residual method) and R-Software (code) in the Bioequivalence 2way cross design.

Regards,
UdayK.

Edit: category changed and two successive posts merged [Ohlbe]
 
ping4santosh
★    

India,
2020-03-11 15:58
(1776 d 13:08 ago)

@ udayk
Posting: # 21250
Views: 5,971
 

 Outlier Test using R-Software and Phoenix WinNonlin software in the BE

Post reply
Hi Uday,

Please let me know more details. I can help you with outliers treatment.

Best,
SKM
 
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2020-03-13 12:21
(1774 d 16:45 ago)

@ udayk
Posting: # 21267
Views: 5,961
 

 Studentized & standardized residuals

Post reply
Hi UdayK,

❝ […] outlier test using Phoenix WinNonlin Software (Studentized residual method) and R-Software (code) in the Bioequivalence 2way cross design.


“Studentized residual” is an ambiguous term. There are two flavors:
  1. Internally studentized (aka “standardized”), where
    \(\widehat{\sigma}^2={1 \over n-m}\sum_{j=1}^n \widehat{\varepsilon\,}_j^{\,2}\)
  2. Externally studentized, where
    \(\widehat{\sigma}_{(i)}^2={1 \over n-m-1}\sum_{\begin{smallmatrix}j = 1\\j \ne i\end{smallmatrix}}^n \widehat{\varepsilon\,}_j^{\,2}\)
#1 is easily accessible in Phoenix WinNonlin (divide the raw residual by the standard error). Assess only the first period. The value of the other period has the same value but with the opposite sign and is therefore, not informative. No idea how to get #2 In WinNonlin. Ask at Certara’s forum.
Both are easily obtained in SAS, R,…

In R for the fixed effects model (untested!):

mod  <- lm(log(PK) ~ sequence + subject%in%sequence + period + treatment, data = yourdata)
res1 <- rstandard(mod) # 1. Internally studentized (standardized)
res2 <- rstudent(mod)  # 2. Externally studentized

Note that #2 is in general more restrictive than #1, i.e., it might be that you detect more outliers.
BTW, which outlier test are you thinking of? See also there and this thread at stackoverflow.
—
Dif-tor heh smusma 🖖🏼 Довге життя Україна! [image]
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
 
ping4santosh
★    

India,
2020-03-14 15:52
(1773 d 13:14 ago)

@ udayk
Posting: # 21269
Views: 5,828
 

 Outlier Test using R-Software: Lund test

Post reply
Hi Uday,

Try Lund test in R package. If you want, I can help you out.

Cheers,
SKM
 
udayk
☆    

India,
2020-03-20 08:05
(1767 d 21:01 ago)

@ ping4santosh
Posting: # 21288
Views: 5,735
 

 Outlier Test using R-Software: Lund test

Post reply
Hi SKM,

Please help us.

Regards,
Uday
 
ping4santosh
★    

India,
2020-03-20 11:07
(1767 d 17:59 ago)

(edited on 2020-03-21 03:37)
@ udayk
Posting: # 21289
Views: 5,633
 

 Outlier Test using R-Software: Lund test

Post reply

❝ Please help us.


Hi Uday: I shall need access to the data and the background info. Cheers, SKM

NB: my email ID is my ID here in Gmail.
 
yjlee168
★★★
avatar
Homepage
Kaohsiung, Taiwan,
2020-03-20 15:45
(1767 d 13:21 ago)

@ udayk
Posting: # 21291
Views: 5,659
 

 bear for outlier detection analysis?

Post reply
Hi udayk,

You may consider bear - an R package if you like. With bear, you can do outlier detection analysis (ODA) for a 2x2x2 BE study. Bear provides the following ODAs:
  1. Test for Normality (Pearson): studentized intra-subject residuals & studentized inter-subject residuals;
  2. Hotelling T2 with Chi-square Test;
  3. Test for Equality of Intra-subject Variabilities between Formulations; and
  4. Cook's distance.
For more info, please see my signature panel.

[image]
—
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
 
xtianbadillo
☆    

Mexico,
2024-12-16 21:44
(35 d 07:22 ago)

@ yjlee168
Posting: # 24317
Views: 810
 

 bear for outlier detection analysis?

Post reply
Good morning,

I was performing some residual analyses while also checking if my bear 2.9.2 was correctly installed on my PC with Windows 10. I reviewed the version 2 validation file and compared it against the residuals, and I discovered that there is an error in my system, although I haven’t yet identified what it is. The residuals I obtain are different from the values reported in the validation. Using the same dataset, I ran it in R to obtain the residuals, and they match those in the validation. Later, I will run an older version I have on Windows 7.

I am sharing the results.


From by Hsin-ya Lee & Yung-jin Lee (09/07/2009) Validation of bear using WNL & SAS
VERSION 2 Page 122 of 162

Intra-subject and Inter-subject Residuals
--------------------------------------------------------------------------
subj Obs Exp Intra Stud_Intra Inter Stud_Inter
1 1 7.398174 7.491068 -0.092894 -1.126022 0.138200 0.723698
2 2 7.300473 7.409175 -0.108702 -1.317638 0.098454 0.515564
3 3 7.636752 7.622009 0.014743 0.178707 0.400081 2.095061
4 4 7.225481 7.311596 -0.086115 -1.043844 -0.096705 -0.506404
5 5 7.233455 7.352791 -0.119336 -1.446539 -0.138354 -0.724503
6 6 7.470794 7.400282 0.070512 0.854712 0.080667 0.422421
7 7 7.404279 7.441728 -0.037449 -0.453938 0.039519 0.206946
8 8 7.569928 7.479503 0.090425 1.096082 0.239110 1.252122
9 9 7.472501 7.445906 0.026595 0.322378 0.047874 0.250697
10 10 7.235619 7.269715 -0.034096 -0.413295 -0.180467 -0.945031
11 11 7.427739 7.288976 0.138763 1.682023 -0.265986 -1.392859
12 12 7.340836 7.235661 0.105175 1.274886 -0.248576 -1.301690
13 13 7.380879 7.311301 0.069578 0.843390 -0.221335 -1.159040
14 14 7.376508 7.413706 -0.037198 -0.450902 0.107516 0.563018

From Bear 2.9.2 Using Single2x2x2_stat_demo Statistical Analysis Only

* Intra-subject and Inter-subject Residuals *
---------------------------------------------
subj Obs Exp Intra Stud_Intra Inter Stud_Inter
1 7.400000 7.490714 -0.090714 -1.100692 0.138571 0.727189
2 7.300000 7.410714 -0.110714 -1.343364 0.098571 0.517279
3 7.640000 7.620714 0.019286 0.234005 0.398571 2.091605
4 7.230000 7.315714 -0.085714 -1.040024 -0.091429 -0.479795
5 7.230000 7.350714 -0.120714 -1.464700 -0.141429 -0.742182
6 7.470000 7.400714 0.069286 0.840686 0.078571 0.412324
7 7.400000 7.440714 -0.040714 -0.494011 0.038571 0.202413
8 7.570000 7.480714 0.089286 1.083358 0.238571 1.251964
9 7.470000 7.445714 0.024286 0.294673 0.048571 0.254891
10 7.240000 7.270714 -0.030714 -0.372675 -0.181429 -0.952092
11 7.430000 7.290714 0.139286 1.690039 -0.261429 -1.371913
12 7.340000 7.235714 0.104286 1.265362 -0.251429 -1.319435
13 7.380000 7.310714 0.069286 0.840686 -0.221429 -1.162003
14 7.380000 7.415714 -0.035714 -0.433343 0.108571 0.569756


From Bear 2.9.2 Using Single2x2x2_stat_demo NCA ->Statistical Analysis Only

---------------------------------------------

* Intra-subject and Inter-subject Residuals *
---------------------------------------------
subj Obs Exp Intra Stud_Intra Inter Stud_Inter
1 7.398174 7.491068 -0.092894 -1.063472 0.138200 0.686507
2 7.300473 7.397679 -0.097206 -1.112839 0.075462 0.374856
3 7.636752 7.622009 0.014743 0.168782 0.400081 1.987400
4 7.225481 7.300100 -0.074618 -0.854246 -0.119697 -0.594594
5 7.233455 7.352792 -0.119336 -1.366186 -0.138354 -0.687270
6 7.470794 7.388785 0.082008 0.938847 0.057674 0.286497
7 7.404279 7.441728 -0.037449 -0.428723 0.039519 0.196312
8 7.569928 7.468007 0.101920 1.166805 0.216118 1.073564
9 7.472501 7.445906 0.026595 0.304466 0.047874 0.237815
10 7.235619 7.258219 -0.022600 -0.258728 -0.203459 -1.010680
11 7.427739 7.288975 0.138764 1.588593 -0.265986 -1.321286
12 7.340836 7.224164 0.116671 1.335678 -0.271569 -1.349016
13 7.380879 7.311301 0.069578 0.796540 -0.221334 -1.099478
14 7.376508 7.482683 -0.106175 -1.215516 0.245470 1.219372


ejemplo <- lm (lnCmax~Secuencia+Sujeto*Secuencia+Tratamiento+Periodo , data=datos)
anova(ejemplo)


valores_predichos <- round(fitted(ejemplo),4)
residuales <- round(residuals(ejemplo),4)
residuales_std <- round(rstandard(ejemplo),4)
residuales_stu <- round(rstudent(ejemplo),4)

valores_observados valores_predichos residuales residuales_std residuales_stu
1 7.4611 7.3682 0.0929 1.1260 1.1400
2 7.3005 7.4092 -0.1087 -1.3176 -1.3641
3 7.4844 7.4991 -0.0147 -0.1787 -0.1713
4 7.2255 7.3116 -0.0861 -1.0438 -1.0481
5 7.3492 7.2299 0.1193 1.4465 1.5242
6 7.4708 7.4003 0.0705 0.8547 0.8444
 
UA Flag
Activity
 Admin contact
23,363 posts in 4,906 threads, 1,678 registered users;
143 visitors (0 registered, 143 guests [including 11 identified bots]).
Forum time: 05:06 CET (Europe/Vienna)

The combination of some data and an aching desire
for an answer does not ensure that a reasonable answer
can be extracted from a given body of data.    John W. Tukey

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5