Add gs_bound_summary() function

[jdblischak]

Closes #468

xref: #498 (comment)

The goal is to create a summary table similar to gsDesign::gsBoundSummary(). This is an initial implementation. Please review and let me know where I need to add extra features and flexibility.

suppressPackageStartupMessages(library("gsDesign"))
x <- gsSurv(alpha = 0.01)
gsBoundSummary(x, alpha = c(0.0125, 0.025))
##     Analysis              Value α=0.01 α=0.0125 α=0.025 Futility
##    IA 1: 33%                  Z 3.2788   3.2153  3.0107  -0.0244
##       N: 234        p (1-sided) 0.0005   0.0007  0.0013   0.5097
##   Events: 71       ~HR at bound 0.4579   0.4649  0.4881   1.0058
##     Month: 8   P(Cross) if HR=1 0.0005   0.0007  0.0013   0.4903
##              P(Cross) if HR=0.6 0.1295   0.1434  0.1947   0.0148
##    IA 2: 67%                  Z 2.8566   2.7838  2.5465   1.2441
##       N: 332        p (1-sided) 0.0021   0.0027  0.0054   0.1067
##  Events: 141       ~HR at bound 0.6181   0.6257  0.6512   0.8110
##    Month: 13   P(Cross) if HR=1 0.0025   0.0031  0.0062   0.8986
##              P(Cross) if HR=0.6 0.5778   0.6061  0.6937   0.0437
##        Final                  Z 2.3695   2.2838  1.9992   2.3695
##       N: 332        p (1-sided) 0.0089   0.0112  0.0228   0.0089
##  Events: 212       ~HR at bound 0.7219   0.7305  0.7596   0.7219
##    Month: 18   P(Cross) if HR=1 0.0093   0.0114  0.0213   0.9907
##              P(Cross) if HR=0.6 0.9000   0.9107  0.9355   0.1000

suppressPackageStartupMessages(library(gsDesign2))

x <- gs_design_ahr(info_frac = c(.25, .75, 1), analysis_time = c(12, 25, 36))
gs_bound_summary(x)
##       Analysis               Value Efficacy Futility
## 1    IA 1: 31%                   Z   3.8728  -1.6993
## 2       N: 439         p (1-sided)   0.0001   0.9554
## 3   Events: 99        ~HR at bound   0.4600   1.4060
## 4    Month: 12    P(Cross) if HR=1   0.0001   0.0446
## 5              P(Cross) if AHR=0.8   0.0025   0.0032
## 6    IA 2: 77%                   Z   2.3100   1.0889
## 7       N: 527         p (1-sided)   0.0104   0.1381
## 8  Events: 248        ~HR at bound   0.7455   0.8707
## 9    Month: 25    P(Cross) if HR=1   0.0105   0.8620
## 10             P(Cross) if AHR=0.7   0.6348   0.0599
## 11       Final                   Z   2.0161   2.0143
## 12      N: 527         p (1-sided)   0.0219   0.0220
## 13 Events: 323        ~HR at bound   0.7990   0.7992
## 14   Month: 36    P(Cross) if HR=1   0.0243   0.9756
## 15             P(Cross) if AHR=0.7   0.9000   0.1001

x <- gs_design_wlr(info_frac = c(.25, .75, 1), analysis_time = c(12, 25, 36))
gs_bound_summary(x)
##       Analysis                Value Efficacy Futility
## 1    IA 1: 25%                    Z   4.2955  -2.0871
## 2       N: 436          p (1-sided)   0.0000   0.9816
## 3   Events: 99         ~HR at bound   0.4213   1.5219
## 4    Month: 12     P(Cross) if HR=1   0.0000   0.0184
## 5              P(Cross) if wAHR=0.8   0.0004   0.0012
## 6    IA 2: 76%                    Z   2.3156   1.1034
## 7       N: 523          p (1-sided)   0.0103   0.1349
## 8  Events: 246         ~HR at bound   0.7442   0.8687
## 9    Month: 25     P(Cross) if HR=1   0.0103   0.8651
## 10             P(Cross) if wAHR=0.7   0.6323   0.0610
## 11       Final                    Z   2.0151   2.0151
## 12      N: 523          p (1-sided)   0.0219   0.0219
## 13 Events: 321         ~HR at bound   0.7985   0.7985
## 14   Month: 36     P(Cross) if HR=1   0.0242   0.9758
## 15             P(Cross) if wAHR=0.7   0.9000   0.1000

[jdblischak]

Codecov coverage report available at https://app.codecov.io/gh/Merck/gsDesign2/pull/515

[LittleBeannie]

Looks great to me! Thanks, @jdblischak!