Compute summaries from network meta-analysis resampling replications
compute_network_alt_stats.RdCompute summaries from network meta-analysis resampling replications
Usage
compute_network_alt_stats(
object,
type = NULL,
quantiles = c(0.025, 0.5, 0.975)
)Examples
# \donttest{
nma_dat <- data.frame(
study = c("S1", "S2", "S3", "S4"),
treatment_1 = c("A", "A", "B", "A"),
treatment_2 = c("B", "C", "C", "B"),
effect = c(0.20, 0.35, 0.10, 0.15),
variance = c(0.04, 0.05, 0.04, 0.06)
)
nma_reps <- replicate_network_bootstrap(
number_bootstraps = 2,
data = nma_dat,
study_id = "study",
treatment_1 = "treatment_1",
treatment_2 = "treatment_2",
effect = "effect",
variance = "variance",
model_type = "wls"
)
compute_network_alt_stats(nma_reps)
#> $summary
#> point_estimate ci_lower ci_upper
#> basic_effects.A 0.0000000 0.0000000 0.0000000
#> basic_effects.B 0.2150000 0.1817500 0.2482500
#> basic_effects.C 0.3150000 0.2817500 0.3482500
#> direct.A_vs_B 0.1800000 0.1800000 0.1800000
#> direct.A_vs_C 0.3500000 0.3500000 0.3500000
#> direct.B_vs_C 0.1000000 0.1000000 0.1000000
#> indirect.A_vs_B 0.2500000 0.2500000 0.2500000
#> indirect.A_vs_C 0.2800000 0.2800000 0.2800000
#> indirect.B_vs_C NA NA NA
#> total.A_vs_B 0.2150000 0.1817500 0.2482500
#> total.A_vs_C 0.3150000 0.2817500 0.3482500
#> total.B_vs_C 0.1000000 0.1000000 0.1000000
#> ranking.mean_effect.C 0.3150000 0.2817500 0.3482500
#> ranking.mean_effect.B 0.2150000 0.1817500 0.2482500
#> ranking.mean_effect.A 0.0000000 0.0000000 0.0000000
#> ranking.std_error.C 0.1839378 0.1594051 0.2084706
#> ranking.std_error.B 0.1835257 0.1563497 0.2107017
#> ranking.std_error.A 0.0000000 0.0000000 0.0000000
#> ranking.mean_rank.C 1.2958750 1.2610812 1.3306688
#> ranking.mean_rank.B 1.8767500 1.8758000 1.8777000
#> ranking.mean_rank.A 2.8273750 2.7935313 2.8612187
#> ranking.median_rank.C 1.0000000 1.0000000 1.0000000
#> ranking.median_rank.B 2.0000000 2.0000000 2.0000000
#> ranking.median_rank.A 3.0000000 3.0000000 3.0000000
#> ranking.sucra.C 85.2062500 83.4665625 86.9459375
#> ranking.sucra.B 56.1625000 56.1150000 56.2100000
#> ranking.sucra.A 8.6312500 6.9390625 10.3234375
#> ranking.sucr_a.C 85.2062500 83.4665625 86.9459375
#> ranking.sucr_a.B 56.1625000 56.1150000 56.2100000
#> ranking.sucr_a.A 8.6312500 6.9390625 10.3234375
#> ranking.p_best.C 0.7397500 0.7347625 0.7447375
#> ranking.p_best.B 0.2285000 0.2144875 0.2425125
#> ranking.p_best.A 0.0317500 0.0127500 0.0507500
#> fit_stats.n_treatments 3.0000000 3.0000000 3.0000000
#> fit_stats.n_edges 2.0000000 2.0000000 2.0000000
#> fit_stats.QE 0.0125000 0.0006250 0.0243750
#> fit_stats.QE_df 2.0000000 2.0000000 2.0000000
#> fit_stats.QE_p 0.9937889 0.9878884 0.9996894
#> fit_stats.logLik 2.5428095 2.5390517 2.5465672
#> fit_stats.Dev -5.0856189 -5.0931344 -5.0781035
#> fit_stats.AIC -1.0856189 -1.0931344 -1.0781035
#> fit_stats.BIC -2.3130302 -2.3205457 -2.3055148
#> fit_stats.AICc 10.9143811 10.9068656 10.9218965
#>
#> $effects_summary
#> effect_type treatment_1 treatment_2 point_estimate ci_lower ci_upper
#> 7 total A B 0.215 0.18175 0.24825
#> 8 total A C 0.315 0.28175 0.34825
#> 9 total B C 0.100 0.10000 0.10000
#> 1 direct A B 0.180 0.18000 0.18000
#> 2 direct A C 0.350 0.35000 0.35000
#> 3 direct B C 0.100 0.10000 0.10000
#> 4 indirect A B 0.250 0.25000 0.25000
#> 5 indirect A C 0.280 0.28000 0.28000
#> 6 indirect B C NA NA NA
#>
#> $basic_effects_summary
#> treatment point_estimate ci_lower ci_upper
#> 1 A 0.000 0.00000 0.00000
#> 2 B 0.215 0.18175 0.24825
#> 3 C 0.315 0.28175 0.34825
#>
#> $quantiles
#> 2.5% 50.0% 97.5%
#> basic_effects.A 0.0000000 0.0000000 0.0000000
#> basic_effects.B 0.1817500 0.2150000 0.2482500
#> basic_effects.C 0.2817500 0.3150000 0.3482500
#> direct.A_vs_B 0.1800000 0.1800000 0.1800000
#> direct.A_vs_C 0.3500000 0.3500000 0.3500000
#> direct.B_vs_C 0.1000000 0.1000000 0.1000000
#> indirect.A_vs_B 0.2500000 0.2500000 0.2500000
#> indirect.A_vs_C 0.2800000 0.2800000 0.2800000
#> indirect.B_vs_C NA NA NA
#> total.A_vs_B 0.1817500 0.2150000 0.2482500
#> total.A_vs_C 0.2817500 0.3150000 0.3482500
#> total.B_vs_C 0.1000000 0.1000000 0.1000000
#> ranking.mean_effect.C 0.2817500 0.3150000 0.3482500
#> ranking.mean_effect.B 0.1817500 0.2150000 0.2482500
#> ranking.mean_effect.A 0.0000000 0.0000000 0.0000000
#> ranking.std_error.C 0.1594051 0.1839378 0.2084706
#> ranking.std_error.B 0.1563497 0.1835257 0.2107017
#> ranking.std_error.A 0.0000000 0.0000000 0.0000000
#> ranking.mean_rank.C 1.2610812 1.2958750 1.3306688
#> ranking.mean_rank.B 1.8758000 1.8767500 1.8777000
#> ranking.mean_rank.A 2.7935313 2.8273750 2.8612187
#> ranking.median_rank.C 1.0000000 1.0000000 1.0000000
#> ranking.median_rank.B 2.0000000 2.0000000 2.0000000
#> ranking.median_rank.A 3.0000000 3.0000000 3.0000000
#> ranking.sucra.C 83.4665625 85.2062500 86.9459375
#> ranking.sucra.B 56.1150000 56.1625000 56.2100000
#> ranking.sucra.A 6.9390625 8.6312500 10.3234375
#> ranking.sucr_a.C 83.4665625 85.2062500 86.9459375
#> ranking.sucr_a.B 56.1150000 56.1625000 56.2100000
#> ranking.sucr_a.A 6.9390625 8.6312500 10.3234375
#> ranking.p_best.C 0.7347625 0.7397500 0.7447375
#> ranking.p_best.B 0.2144875 0.2285000 0.2425125
#> ranking.p_best.A 0.0127500 0.0317500 0.0507500
#> fit_stats.n_treatments 3.0000000 3.0000000 3.0000000
#> fit_stats.n_edges 2.0000000 2.0000000 2.0000000
#> fit_stats.QE 0.0006250 0.0125000 0.0243750
#> fit_stats.QE_df 2.0000000 2.0000000 2.0000000
#> fit_stats.QE_p 0.9878884 0.9937889 0.9996894
#> fit_stats.logLik 2.5390517 2.5428095 2.5465672
#> fit_stats.Dev -5.0931344 -5.0856189 -5.0781035
#> fit_stats.AIC -1.0931344 -1.0856189 -1.0781035
#> fit_stats.BIC -2.3205457 -2.3130302 -2.3055148
#> fit_stats.AICc 10.9068656 10.9143811 10.9218965
#>
#> $number_repl
#> [1] 2
#>
#> $number_success
#> [1] 2
#>
#> $number_failed
#> [1] 0
#>
#> $robust_type
#> [1] "bootstrap"
#>
#> $boot_repl
#> basic_effects.A basic_effects.B basic_effects.C direct.A_vs_B direct.A_vs_C
#> 1 0 0.18 0.28 0.18 NA
#> 2 0 0.25 0.35 NA 0.35
#> direct.B_vs_C indirect.A_vs_B indirect.A_vs_C indirect.B_vs_C total.A_vs_B
#> 1 0.1 NA 0.28 NA 0.18
#> 2 0.1 0.25 NA NA 0.25
#> total.A_vs_C total.B_vs_C ranking.mean_effect.C ranking.mean_effect.B
#> 1 0.28 0.1 0.28 0.18
#> 2 0.35 0.1 0.35 0.25
#> ranking.mean_effect.A ranking.std_error.C ranking.std_error.B
#> 1 0 0.2097618 0.1549193
#> 2 0 0.1581139 0.2121320
#> ranking.std_error.A ranking.mean_rank.C ranking.mean_rank.B
#> 1 0 1.33250 1.87575
#> 2 0 1.25925 1.87775
#> ranking.mean_rank.A ranking.median_rank.C ranking.median_rank.B
#> 1 2.79175 1 2
#> 2 2.86300 1 2
#> ranking.median_rank.A ranking.sucra.C ranking.sucra.B ranking.sucra.A
#> 1 3 83.3750 56.2125 10.4125
#> 2 3 87.0375 56.1125 6.8500
#> ranking.sucr_a.C ranking.sucr_a.B ranking.sucr_a.A ranking.p_best.C
#> 1 83.3750 56.2125 10.4125 0.7345
#> 2 87.0375 56.1125 6.8500 0.7450
#> ranking.p_best.B ranking.p_best.A fit_stats.n_treatments fit_stats.n_edges
#> 1 0.21375 0.05175 3 2
#> 2 0.24325 0.01175 3 2
#> fit_stats.QE fit_stats.QE_df fit_stats.QE_p fit_stats.logLik fit_stats.Dev
#> 1 2.500000e-02 2 0.9875778 2.546765 -5.093530
#> 2 1.617781e-31 2 1.0000000 2.538854 -5.077708
#> fit_stats.AIC fit_stats.BIC fit_stats.AICc
#> 1 -1.093530 -2.320941 10.90647
#> 2 -1.077708 -2.305119 10.92229
#>
#> $jack_repl
#> NULL
#>
#> $failures
#> list()
#>
#> $model_information
#> $model_information$object
#> $model_information$object$structure
#> [1] "wls"
#>
#> $model_information$object$estimation_method
#> [1] "WLS"
#>
#> $model_information$object$formula
#> network_meta(data = data, study_id = study_id, treatment_1 = treatment_1,
#> treatment_2 = treatment_2, effect = effect, variance = variance,
#> model_type = "wls")
#>
#> $model_information$object$fit_stats
#> model_type heterogeneity n_treatments n_edges estimator QE QE_df QE_p
#> 1 <NA> <NA> 3 2 WLS 0.025 2 0.9875778
#> logLik Dev AIC BIC AICc
#> 1 2.546765 -5.09353 -1.09353 -2.320941 10.90647
#>
#>
#> $model_information$dimensions
#> $model_information$dimensions$number_effectsize
#> [1] 4
#>
#> $model_information$dimensions$number_fixed
#> [1] 3
#>
#> $model_information$dimensions$number_random
#> [1] NA
#>
#>
#> $model_information$beta
#> attribute
#> 1 basic_effects.A
#> 2 basic_effects.B
#> 3 basic_effects.C
#>
#> $model_information$analysis_type
#> [1] "network_meta"
#>
#>
#> attr(,"class")
#> [1] "nma_boot" "mars_boot"
# }