This function will calculate the summary statistics for a specific response outcome scenario.
Usage
sumTable(
thisResp,
TotalSample,
cut_B,
cut_W,
parX,
YPri = c(0.5, 0.5),
Round = 2
)Arguments
- thisResp
number of responses
- TotalSample
sample size
- cut_B
a meaningful improvement threshold: at least cut_B (e.g. 15%) improvement
- cut_W
a poor improvement threshold: at most cut_W (e.g. 5%) improvement
- parX
non-negative parameters of the beta distribution of the control (posterior)
- YPri
non-negative parameters of the beta prior of the treatment, default Beta(0.5,0.5)
- Round
Rounding of the output statistics
Examples
sumTable(2, 25, parX = c(1, 52), cut_B = 0.2, cut_W = 0.05, YPri = c(1, 1))
#> summaries
#> # resp 2.00
#> obs ORR [%] 8.00
#> mode [%] 6.48
#> CI lower [%] 0.71
#> CI upper [%] 20.68
#> prob.go [%] 5.87
#> prob.nogo [%] 26.59
# for multiple response scenarios (e.g. 0 to 8 responses out of 25)
summaries <- do.call(
cbind,
lapply(c(0:8),
sumTable,
TotalSample = 25,
parX = c(1, 52),
cut_B = 0.2,
cut_W = 0.05
)
)
summaries
#> summaries summaries summaries summaries summaries summaries
#> # resp 0.00 1.00 2.00 3.00 4.00 5.00
#> obs ORR [%] 0.00 4.00 8.00 12.00 16.00 20.00
#> mode [%] 0.00 1.02 4.80 8.83 12.92 17.04
#> CI lower [%] -4.55 -2.42 -0.24 2.03 4.55 7.27
#> CI upper [%] 17.51 13.05 18.83 24.07 28.97 33.64
#> prob.go [%] 0.05 0.74 3.79 11.58 25.23 43.17
#> prob.nogo [%] 93.39 66.49 36.46 16.09 6.00 1.98
#> summaries summaries summaries
#> # resp 6.00 7.00 8.00
#> obs ORR [%] 24.00 28.00 32.00
#> mode [%] 21.18 25.32 29.47
#> CI lower [%] 10.17 13.20 16.37
#> CI upper [%] 38.13 42.46 46.67
#> prob.go [%] 61.70 77.26 88.13
#> prob.nogo [%] 0.59 0.17 0.04