Update script to generate MBdelayed.rda

[jdblischak]

I updated data-raw/DATASET.R to use the new snakecase function names and lowercase column names. However, I was unable to reproduce the existing object data/MBdelayed.rda. My updated version has no events.

Does anyone have any ideas on how to fix the script so that it can reproduce the existing object?

Here are some diagnostics:

# current MBdelayed
summary(MBdelayed)
##       tte               event         Stratum           Treatment        
##  Min.   : 0.09171   Min.   :0.000   Length:200         Length:200        
##  1st Qu.: 5.85919   1st Qu.:0.000   Class :character   Class :character  
##  Median :14.49925   Median :1.000   Mode  :character   Mode  :character  
##  Mean   :16.79899   Mean   :0.705                                        
##  3rd Qu.:27.37013   3rd Qu.:1.000                                        
##  Max.   :35.42239   Max.   :1.000           
str(MBdelayed)
## gropd_df [200 × 4] (S3: grouped_df/tbl_df/tbl/data.frame)
##  $ tte      : num [1:200] 4.68 6.2 35.42 32.74 5.38 ...
##  $ event    : num [1:200] 1 1 1 1 1 1 1 1 1 1 ...
##  $ Stratum  : chr [1:200] "All" "All" "All" "All" ...
##  $ Treatment: chr [1:200] "Experimental" "Control" "Control" "Experimental" ...
##  - attr(*, "groups")= tibble [2 × 3] (S3: tbl_df/tbl/data.frame)
##   ..$ Stratum  : chr [1:2] "All" "All"
##   ..$ Treatment: chr [1:2] "Control" "Experimental"
##   ..$ .rows    :List of 2
##   .. ..$ : int [1:100] 2 3 7 8 9 10 14 15 17 18 ...
##   .. ..$ : int [1:100] 1 4 5 6 11 12 13 16 19 20 ...
##   ..- attr(*, ".drop")= logi TRUE
table(MBdelayed[, c("event", "Stratum", "Treatment")])
## , , Treatment = Control
## 
##      Stratum
## event All
##     0  20
##     1  80
## 
## , , Treatment = Experimental
## 
##      Stratum
## event All
##     0  39
##     1  61

# latest MBdelayed
summary(MBdelayed)
##       tte            event     stratum           treatment        
##  Min.   :24.00   Min.   :0   Length:200         Length:200        
##  1st Qu.:27.11   1st Qu.:0   Class :character   Class :character  
##  Median :29.76   Median :0   Mode  :character   Mode  :character  
##  Mean   :29.88   Mean   :0                                        
##  3rd Qu.:32.71   3rd Qu.:0                                        
##  Max.   :36.30   Max.   :0                   
str(MBdelayed)
## 'data.frame':	200 obs. of  4 variables:
##  $ tte      : num  36.3 36.3 36.1 36.1 36 ...
##  $ event    : num  0 0 0 0 0 0 0 0 0 0 ...
##  $ stratum  : chr  "All" "All" "All" "All" ...
##  $ treatment: chr  "experimental" "control" "control" "experimental" ...
table(MBdelayed[, c("event", "stratum", "treatment")])
## , , treatment = control
## 
##      stratum
## event All
##     0 100
## 
## , , treatment = experimental
## 
##      stratum
## event All
##     0 100

[nanxstats]

I can get proper events using an older GitHub release version of simtrial:

remotes::install_github("Merck/simtrial@v0.2.2")

then run data-raw/DATASET.R. This might help the triage the specific commit(s) that bring the changed behavior. I guess It won't be too difficult to figure out - possibly related to the earlier refactoring effort this Spring.

[jdblischak]

Confirmed. I was able to run the code from the tag v0.2.2:

table(MBdelayed[, c("event", "Stratum", "Treatment")])
, , Treatment = Control

     Stratum
event All
    0  23
    1  77

, , Treatment = Experimental

     Stratum
event All
    0  38
    1  62

However, the object data/MBdelayed.rda was still modified. And note that the numbers don't match the current data set. I suspect that potentially the seed wasn't set the last time it was updated

[jdblischak]

I also added code to fail early if the input to sim_pw_surv() has mismatched treatment names

[nanxstats]

Very nice - great to have this dataset updated and the functions safe guarded.