conroylau
commented
4 years ago Done! Integrated the function of generating multiple CIs into the main function invertci.
Based on whether alpha is a scalar or vector, print and summary will print the correct information.
Some sample outputs for print and summary below (I have omitted some rows below for brevity):
Vector-valued alpha
print
< Significance level: 0.01 >
Total number of iterations: 20.
Confidence interval: [0.35977, 0.6582].
--------------------------------------
< Significance level: 0.05 >
Total number of iterations: 20.
Confidence interval: [0.36289, 0.65273].summary
Call:
invertci(f = dkqs, farg = farg, alpha = c(0.01, 0.05), lb0 = 0,
lb1 = 0.4, ub0 = 1, ub1 = 0.6, tol = 0.001, max_iter = 50,
df_ci = NULL, progress = TRUE)
Significance levels considered: 0.01 0.05
< Confidence interval for alpha = 0.01 >
=== Iterations in constructing upper bound:
Iteration Test point Lower bound Upper bound p-value Decision
Left end-pt. 0.60000 0.60000 NA 0.77780 Do not reject
Right end-pt. 1.00000 NA 1.00000 0.00000 Reject
1 0.80000 0.60000 1.00000 0.00000 Reject
2 0.70000 0.60000 0.80000 0.00000 Reject
......
8 0.65781 0.65625 0.65938 0.00660 Do not reject
9 0.65859 0.65781 0.65938 0.00380 Reject
>>> Reached maximum number of iterations
=== Iterations in constructing lower bound:
Iteration Test point Lower bound Upper bound p-value Decision
Left end-pt. 0.00000 0.00000 NA 0.00000 Reject
Right end-pt. 0.40000 NA 0.40000 0.78790 Do not reject
1 0.20000 0.00000 0.40000 0.00000 Reject
2 0.30000 0.20000 0.40000 0.00000 Reject
......
8 0.36094 0.35938 0.36250 0.00790 Do not reject
9 0.36016 0.35938 0.36094 0.00610 Do not reject
>>> Reached maximum number of iterations
< Confidence interval for alpha = 0.05 >
=== Iterations in constructing upper bound:
Iteration Test point Lower bound Upper bound p-value Decision
Left end-pt. 0.60000 0.60000 NA 0.77780 Do not reject
Right end-pt. 1.00000 NA 1.00000 0.00000 Reject
1 0.80000 0.60000 1.00000 0.00000 Reject
2 0.70000 0.60000 0.80000 0.00000 Reject
......
8 0.65156 0.65000 0.65312 0.03000 Do not reject
9 0.65234 0.65156 0.65312 0.02570 Do not reject
>>> Reached maximum number of iterations
=== Iterations in constructing lower bound:
Iteration Test point Lower bound Upper bound p-value Decision
Left end-pt. 0.00000 0.00000 NA 0.00000 Reject
Right end-pt. 0.40000 NA 0.40000 0.78790 Do not reject
1 0.20000 0.00000 0.40000 0.00000 Reject
2 0.30000 0.20000 0.40000 0.00000 Reject
......
8 0.36406 0.36250 0.36562 0.03300 Do not reject
9 0.36328 0.36250 0.36406 0.02780 Do not reject Scalar-valued alpha
print
< Significance level: 0.01 >
Total number of iterations: 20.
Confidence interval: [0.35977, 0.6582].summary
Call:
invertci(f = dkqs, farg = farg, alpha = c(0.01), lb0 = 0,
lb1 = 0.4, ub0 = 1, ub1 = 0.6, tol = 0.001, max_iter = 50,
df_ci = NULL, progress = TRUE)
Significance level: 0.01
< Confidence interval for alpha = 0.01 >
=== Iterations in constructing upper bound:
Iteration Test point Lower bound Upper bound p-value Decision
Left end-pt. 0.60000 0.60000 NA 0.77780 Do not reject
Right end-pt. 1.00000 NA 1.00000 0.00000 Reject
1 0.80000 0.60000 1.00000 0.00000 Reject
2 0.70000 0.60000 0.80000 0.00000 Reject
......
8 0.65781 0.65625 0.65938 0.00660 Do not reject
9 0.65859 0.65781 0.65938 0.00380 Reject
>>> Reached maximum number of iterations
=== Iterations in constructing lower bound:
Iteration Test point Lower bound Upper bound p-value Decision
Left end-pt. 0.00000 0.00000 NA 0.00000 Reject
Right end-pt. 0.40000 NA 0.40000 0.78790 Do not reject
1 0.20000 0.00000 0.40000 0.00000 Reject
2 0.30000 0.20000 0.40000 0.00000 Reject
......
8 0.36094 0.35938 0.36250 0.00790 Do not reject
9 0.36016 0.35938 0.36094 0.00610 Do not reject
>>> Reached maximum number of iterations
Instead of having a separate function, just let
qpcibe a driver that does something different depending on whether alpha is a vector or a scalar.