Bioequivalence and Bioavailability Forum • A small point for the code

A small point for the code [GxP / QC / QA]

posted by Helmut – Vienna, Austria, 2022-06-01 16:11 (688 d 06:30 ago) – Posting: # 23038
Views: 3,351

Hi Sameep,

corrected in the meantime; use:

    nseq       <- length(unique(data$Sequence))

    seqs       <- character()

    for (subj in seq_along(subjects)) {

      seqs <- c(seqs, unique(as.character(

                               data$Sequence[data$Subject ==

                                             subjects[subj]])))

      if (length(unique(seqs)) >= nseq) break

    }

    ow         <- options("digits") # Save original options

    options(digits = 12) # More digits for anova

    on.exit(ow)          # Ensure that options are reset if an error occurs

    for (i in subj:n) {

      tmp               <- head(data, i * 2)

      res$subjects[i-1] <- length(unique(tmp$Subject))

      # Acc. to GLs: Nested model is bogus; Subject works as well

      m                 <- lm(log(Cmax) ~ Sequence +

                                          Subject %in% Sequence +

                                          Period +

                                          Treatment, data = tmp)

      res$PE[i-1]       <- exp(coef(m)[["TreatmentT"]])

      if (res$subjects[i-1] > nseq) {

        res[i-1, 3:4]  <- as.numeric(exp(confint(m, "TreatmentT",

                                                 level = 1 - 2 * alpha)))

        res$MSE[i-1]   <- anova(m)[["Residuals", "Mean Sq"]]



        res$CV[i-1]    <- 100 * mse2CV(res$MSE[i-1])

        res$power[i-1] <- suppressMessages( # if unbalanced

                            power.TOST(CV = res$CV[i-1] / 100,

                                       theta0 = res$PE[i-1],

                                       n = length(unique(tmp$Subject))))

        if (res$lower[i-1] >= theta1 & res$upper[i-1] <= theta2)

          res$BE[i-1] <- TRUE

      }

    }

    res            <- res[complete.cases(res), ] # Remove NAs

—
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

Complete thread:

Another two... Ohlbe 2021-09-16 23:41
- Another two... ElMaestro 2021-09-17 08:42
  - Dates Ohlbe 2021-09-17 10:15
  - Some more info Ohlbe 2021-09-17 20:28
    - EU Article 31 referral started Ohlbe 2022-01-24 17:47
- Cheating Helmut 2021-09-17 16:50
  - Increased variability Ohlbe 2021-09-17 17:56
    - Increased variability ElMaestro 2021-09-17 18:51
      - Visualization Helmut 2021-09-18 15:26
- Another two... jag009 2021-09-19 09:18
- Another two... ElMaestro 2022-05-24 12:00
  - Blind monitors or greedy sponsors? Helmut 2022-05-24 12:51
    - Blind monitors or greedy sponsors? Ohlbe 2022-05-24 14:21
      - Blind monitors or greedy sponsors? Helmut 2022-05-24 16:35
    - Blind monitors or greedy sponsors? ElMaestro 2022-05-25 08:11
      - (Cumulative) T/R-ratio vs. time Helmut 2022-05-25 09:19
        
        Sponsors and CRO selection Ohlbe 2022-05-25 10:53
        
        I still think that T/R-ratios are useful Helmut 2022-05-25 12:04
        
        T/R-ratios are useful Ohlbe 2022-05-25 14:40
        
        T/R-ratios are useful Helmut 2022-05-25 15:35
        Bust the Buster Helmut 2022-05-26 15:32
        
        complicate the assessor's life mittyri 2022-05-26 18:28
        
        complicate the assessor's life PharmCat 2022-05-27 09:57
        complicate the assessor's life Helmut 2022-05-27 10:11
        
        Thanks for Busting the Buster sameep 2022-05-31 08:34
        
        Thanks for Busting the Buster Helmut 2022-05-31 15:22
        
        A small point for the code sameep 2022-06-01 13:17
        
        A small point for the codeHelmut 2022-06-01 14:11