Chapter 2 Mean Difference
2.1 はじめに
LeVasseur et al. (2021) は、ダイエットに関する介入研究のネットワークメタ分析である。介入方法としては、運動、栄養、運動および栄養など多岐にわたる。
今回のデータでは、３群RCTであっても、有酸素運動（AER）が２群ある研究がいくつか含まれている。
2.2 データ
Contrast based を読み込む。
library(readxl)
library(esc)
dfPrepostCont1 <- read_excel("data/MD.xlsx",
sheet = "data4r_prepost_cont")
dfPrepostCont2 <- as.data.frame(
esc_mean_gain(pre1mean = dfPrepostCont1$y1b,
pre1sd = dfPrepostCont1$sd1b,
post1mean = dfPrepostCont1$y1e,
post1sd = dfPrepostCont1$sd1e,
grp1n = dfPrepostCont1$N1,
pre2mean = dfPrepostCont1$y2b,
pre2sd = dfPrepostCont1$sd2b,
post2mean = dfPrepostCont1$y2e,
post2sd = dfPrepostCont1$sd2e,
grp2n = dfPrepostCont1$N2,
study = dfPrepostCont1$Study,
es.type = "d"))Arm based を読み込む。
データは、「介入前後の体重の差と標準誤差」と「介入前後の体重と標準偏差」の２種類がある。まず前者を読み込み、後者は esc パッケージを使用し前者に変換して結合する。
library(readxl)
library(esc)
dfMeanArm <- read_excel("data/MD.xlsx",
sheet = "data4r_mean_arm")
df_prepost <- read_excel("data/MD.xlsx",
sheet = "data4r_prepost_arm")
SP_calc <- as.data.frame(esc_mean_sd(grp1m = df_prepost$weight1,
grp1sd = df_prepost$sd1,
grp1n = df_prepost$N,
grp2m = df_prepost$weight2,
grp2sd = df_prepost$sd2,
grp2n = df_prepost$N,
study = df_prepost$Study,
es.type = "d"))
dfTemp2 <- data.frame(Diff = SP_calc$es,
StdErr = SP_calc$se,
N = SP_calc$sample.size,
Treatment = df_prepost$treatment,
Study = SP_calc$study)
df <- rbind(dfMeanArm,dfTemp2)さらに、group 列を作成する。
df$group <- "Combination Interventions"
df[df$Treatment == "SC",]$group <- "Control"
df[df$Treatment == "AER",]$group <- "Exercise Interventions"
df[df$Treatment == "RES",]$group <- "Exercise Interventions"
df[df$Treatment == "AER+RES",]$group <- "Exercise Interventions"
df[df$Treatment == "LOCAL",]$group <- "Dietary Interventions"
df[df$Treatment == "LOFAT",]$group <- "Dietary Interventions"
df[df$Treatment == "LOCARB",]$group <- "Dietary Interventions"
df[df$Treatment == "NCI",]$group <- "Dietary Interventions"
df[df$Treatment == "PHYTO/PLANT",]$group <- "Dietary Interventions"
df[df$Treatment == "MEDIT",]$group <- "Dietary Interventions"
df[df$Treatment == "LOCAL+LOCARB",]$group <- "Dietary Interventions"
df[df$Treatment == "LOFAT+PYTO",]$group <- "Dietary Interventions"2.3 netmeta 頻度論
2.3.1 データ
arm-based データのため、contrast-based (pairwise) データに変換する。
library(netmeta)
dfCB <- pairwise(
treat = Treatment,
n = N,
mean = Diff,
sd = StdErr,
studlab = Study,
reference.group = "SC",
data = df)エラー: Identical treatments for the following studies: 'Demark-Wahnefried et al 2014' - 'Dolan et al 2016' - 'Harrigan et al 2016' - 'Segal et al 2001' Please check dataset.やはり、同じ治療名なのでエラーが出た。
df[6,]$Treatment <- "LOWCAL+AER/RES2"
df[34,]$Treatment <- "AER2"
df[39,]$Treatment <- "AER2"
df[42,]$Treatment <- "LOWCAL+LOWFAT+AER2"
df[95,]$Treatment <- "AER2"
df[143,]$Treatment <- "AER2"
#df <- df[-c(6,34,39,42),]arm based 形式のデータフレームから、pairwise() 関数を用いて contrast-based のデータフレームを作成する。
library(netmeta)
dfCB <- pairwise(
treat = Treatment,
n = N,
TE = Diff,
seTE = StdErr,
studlab = Study,
reference.group = "SC",
data = df)## Number of studies: k = 88
## Number of pairwise comparisons: m = 88
## Number of treatments: n = 22
## Number of designs: d = 30
## Number of subnetworks: 1サブネットワーク数が１であることを確認した。サブネットワークが２以上であると、ネットワークが二つ以上あることになり、すべての治療の比較ができなくなる。
なお、サブネットワークがある場合、subset = nc1$subnet == 2 で指定する。
2.3.2 Network plot
グラフを描く。
netgraph(netmetaLeV,
plastic = FALSE, # 3Dではなくする
points = TRUE, # ノードを表示する
thickness = "number.of.studies", # 線の太さを研究数にする
multiarm = TRUE)
ネットワークグラフの３次元版。環境によっては RStudio が落ちることがある。
2.3.3 要約
要約。データが多いため出力は省略。最後に tau^2、I^2、Q といった指標が表示される。これらは NMA ではなくメタ分析の指標である。tau^2 は、「真の効果の分散」を表す。I^2 は、異質性の指標であり、25%以下であれば低い、50%で中程度、75%でかなりの異質性と判断される。Q は I^2 を計算するために用いられる。
2.3.4 一貫性の評価
## Q statistics to assess homogeneity / consistency
##
## Q df p-value
## Total 179.08 56 < 0.0001
## Within designs 157.24 42 < 0.0001
## Between designs 21.84 14 0.0819
##
## Design-specific decomposition of within-designs Q statistic
##
## Design Q df p-value
## SC vs AER 9.93 8 0.2701
## SC vs AER+RES 43.34 12 < 0.0001
## SC vs LOWCAL+AER 36.67 4 < 0.0001
## SC vs LOWCAL+AER/RES 6.76 3 0.0798
## SC vs LOWFAT/PHYTO+AER/RES 17.03 1 < 0.0001
## SC vs LOWFAT+AER 32.73 4 < 0.0001
## SC vs MEDIT 0.12 1 0.7260
## SC vs RES 0.19 3 0.9788
## SC vs AER vs AER2 10.46 6 0.1066
##
## Between-designs Q statistic after detaching of single designs
##
## Detached design Q df p-value
## LOWFAT+PHYTO vs NCI 20.81 13 0.0769
## NCI vs PHYTO/PLANT 20.81 13 0.0769
## SC vs AER 21.12 13 0.0706
## SC vs AER+RES 21.66 13 0.0609
## SC vs LOWCAL 21.55 13 0.0628
## SC vs LOWCAL+AER/RES 19.09 13 0.1202
## SC vs LOWCAL+LOWFAT+AER 21.80 13 0.0585
## SC vs LOWFAT 15.66 13 0.2682
## SC vs LOWFAT/PHYTO+AER/RES 21.56 13 0.0626
## SC vs LOWFAT+PHYTO 20.81 13 0.0769
## SC vs MEDIT 20.74 13 0.0782
## SC vs PHYTO/PLANT 20.81 13 0.0769
## SC vs RES 21.63 13 0.0614
## SC vs AER vs AER+RES 14.53 12 0.2679
## SC vs AER vs RES 21.63 12 0.0419
## SC vs AER+RES vs LOWFAT/PHYTO+AER/RES 21.56 12 0.0428
## SC vs LOWCAL vs LOWFAT 17.50 12 0.1316
## SC vs LOWFAT vs MEDIT 16.11 12 0.1864
##
## Q statistic to assess consistency under the assumption of
## a full design-by-treatment interaction random effects model
##
## Q df p-value tau.within tau2.within
## Between designs 11.36 14 0.6573 0.7808 0.6096次の図はあまり必要ないかもしれない。直接か間接かを示す（結果は示さず）。
## Separate indirect from direct evidence (SIDE) using back-calculation method
##
## Common effects model:
##
## comparison k prop nma direct indir. Diff z p-value
## AER vs AER+RES 1 0.08 -0.141 -0.745 -0.089 -0.656 -0.97 0.3322
## AER vs AER2 4 0.84 0.190 0.184 0.226 -0.042 -0.07 0.9468
## AER vs LOWCAL 0 0 2.191 . 2.191 . . .
## AER vs LOWCAL+AER 0 0 1.020 . 1.020 . . .
## AER vs LOWCAL+AER/RES 0 0 0.728 . 0.728 . . .
## AER vs LOWCAL+AER/RES2 0 0 0.370 . 0.370 . . .
## AER vs LOWCAL+LOWFAT 0 0 -0.250 . -0.250 . . .
## AER vs LOWCAL+LOWFAT+AER 0 0 3.547 . 3.547 . . .
## AER vs LOWCAL+LOWFAT+AER2 0 0 2.824 . 2.824 . . .
## AER vs LOWCARB 0 0 -0.736 . -0.736 . . .
## AER vs LOWCARB+AER 0 0 10.396 . 10.396 . . .
## AER vs LOWFAT 0 0 -0.336 . -0.336 . . .
## AER vs LOWFAT/PHYTO+AER/RES 0 0 -0.117 . -0.117 . . .
## AER vs LOWFAT+AER 0 0 -0.235 . -0.235 . . .
## AER vs LOWFAT+PHYTO 0 0 -0.832 . -0.832 . . .
## AER vs MEDIT 0 0 -0.459 . -0.459 . . .
## AER vs MEDIT+AER 0 0 -0.162 . -0.162 . . .
## AER vs NCI 0 0 -0.844 . -0.844 . . .
## AER vs PHYTO/PLANT 0 0 -0.147 . -0.147 . . .
## AER vs RES 1 0.06 -0.337 -0.600 -0.321 -0.279 -0.40 0.6901
## AER vs SC 15 0.96 -0.204 -0.198 -0.339 0.141 0.19 0.8463
## AER+RES vs AER2 0 0 0.331 . 0.331 . . .
## AER+RES vs LOWCAL 0 0 2.332 . 2.332 . . .
## AER+RES vs LOWCAL+AER 0 0 1.161 . 1.161 . . .
## AER+RES vs LOWCAL+AER/RES 0 0 0.868 . 0.868 . . .
## AER+RES vs LOWCAL+AER/RES2 0 0 0.510 . 0.510 . . .
## AER+RES vs LOWCAL+LOWFAT 0 0 -0.110 . -0.110 . . .
## AER+RES vs LOWCAL+LOWFAT+AER 0 0 3.687 . 3.687 . . .
## AER+RES vs LOWCAL+LOWFAT+AER2 0 0 2.965 . 2.965 . . .
## AER+RES vs LOWCARB 0 0 -0.596 . -0.596 . . .
## AER+RES vs LOWCARB+AER 0 0 10.537 . 10.537 . . .
## AER+RES vs LOWFAT 0 0 -0.196 . -0.196 . . .
## AER+RES vs LOWFAT/PHYTO+AER/RES 1 0.17 0.024 -0.137 0.058 -0.195 -0.46 0.6423
## AER+RES vs LOWFAT+AER 0 0 -0.095 . -0.095 . . .
## AER+RES vs LOWFAT+PHYTO 0 0 -0.692 . -0.692 . . .
## AER+RES vs MEDIT 0 0 -0.319 . -0.319 . . .
## AER+RES vs MEDIT+AER 0 0 -0.021 . -0.021 . . .
## AER+RES vs NCI 0 0 -0.704 . -0.704 . . .
## AER+RES vs PHYTO/PLANT 0 0 -0.006 . -0.006 . . .
## AER+RES vs RES 0 0 -0.197 . -0.197 . . .
## AER+RES vs SC 15 0.95 -0.063 -0.083 0.337 -0.421 -0.74 0.4616
## AER2 vs LOWCAL 0 0 2.001 . 2.001 . . .
## AER2 vs LOWCAL+AER 0 0 0.830 . 0.830 . . .
## AER2 vs LOWCAL+AER/RES 0 0 0.537 . 0.537 . . .
## AER2 vs LOWCAL+AER/RES2 0 0 0.180 . 0.180 . . .
## AER2 vs LOWCAL+LOWFAT 0 0 -0.441 . -0.441 . . .
## AER2 vs LOWCAL+LOWFAT+AER 0 0 3.356 . 3.356 . . .
## AER2 vs LOWCAL+LOWFAT+AER2 0 0 2.634 . 2.634 . . .
## AER2 vs LOWCARB 0 0 -0.927 . -0.927 . . .
## AER2 vs LOWCARB+AER 0 0 10.206 . 10.206 . . .
## AER2 vs LOWFAT 0 0 -0.527 . -0.527 . . .
## AER2 vs LOWFAT/PHYTO+AER/RES 0 0 -0.307 . -0.307 . . .
## AER2 vs LOWFAT+AER 0 0 -0.425 . -0.425 . . .
## AER2 vs LOWFAT+PHYTO 0 0 -1.023 . -1.023 . . .
## AER2 vs MEDIT 0 0 -0.650 . -0.650 . . .
## AER2 vs MEDIT+AER 0 0 -0.352 . -0.352 . . .
## AER2 vs NCI 0 0 -1.034 . -1.034 . . .
## AER2 vs PHYTO/PLANT 0 0 -0.337 . -0.337 . . .
## AER2 vs RES 0 0 -0.527 . -0.527 . . .
## AER2 vs SC 4 0.78 -0.394 -0.392 -0.401 0.009 0.02 0.9873
## LOWCAL vs LOWCAL+AER 0 0 -1.171 . -1.171 . . .
## LOWCAL vs LOWCAL+AER/RES 0 0 -1.464 . -1.464 . . .
## LOWCAL vs LOWCAL+AER/RES2 0 0 -1.822 . -1.822 . . .
## LOWCAL vs LOWCAL+LOWFAT 0 0 -2.442 . -2.442 . . .
## LOWCAL vs LOWCAL+LOWFAT+AER 0 0 1.355 . 1.355 . . .
## LOWCAL vs LOWCAL+LOWFAT+AER2 0 0 0.633 . 0.633 . . .
## LOWCAL vs LOWCARB 0 0 -2.928 . -2.928 . . .
## LOWCAL vs LOWCARB+AER 0 0 8.205 . 8.205 . . .
## LOWCAL vs LOWFAT 1 0.67 -2.528 -1.400 -4.801 3.401 2.05 0.0404
## LOWCAL vs LOWFAT/PHYTO+AER/RES 0 0 -2.308 . -2.308 . . .
## LOWCAL vs LOWFAT+AER 0 0 -2.427 . -2.427 . . .
## LOWCAL vs LOWFAT+PHYTO 0 0 -3.024 . -3.024 . . .
## LOWCAL vs MEDIT 0 0 -2.651 . -2.651 . . .
## LOWCAL vs MEDIT+AER 0 0 -2.353 . -2.353 . . .
## LOWCAL vs NCI 0 0 -3.036 . -3.036 . . .
## LOWCAL vs PHYTO/PLANT 0 0 -2.338 . -2.338 . . .
## LOWCAL vs RES 0 0 -2.528 . -2.528 . . .
## LOWCAL vs SC 2 0.70 -2.395 -3.447 0.108 -3.555 -2.08 0.0377
## LOWCAL+AER vs LOWCAL+AER/RES 0 0 -0.293 . -0.293 . . .
## LOWCAL+AER vs LOWCAL+AER/RES2 0 0 -0.651 . -0.651 . . .
## LOWCAL+AER vs LOWCAL+LOWFAT 0 0 -1.271 . -1.271 . . .
## LOWCAL+AER vs LOWCAL+LOWFAT+AER 0 0 2.526 . 2.526 . . .
## LOWCAL+AER vs LOWCAL+LOWFAT+AER2 0 0 1.804 . 1.804 . . .
## LOWCAL+AER vs LOWCARB 0 0 -1.757 . -1.757 . . .
## LOWCAL+AER vs LOWCARB+AER 0 0 9.376 . 9.376 . . .
## LOWCAL+AER vs LOWFAT 0 0 -1.357 . -1.357 . . .
## LOWCAL+AER vs LOWFAT/PHYTO+AER/RES 0 0 -1.137 . -1.137 . . .
## LOWCAL+AER vs LOWFAT+AER 0 0 -1.256 . -1.256 . . .
## LOWCAL+AER vs LOWFAT+PHYTO 0 0 -1.853 . -1.853 . . .
## LOWCAL+AER vs MEDIT 0 0 -1.480 . -1.480 . . .
## LOWCAL+AER vs MEDIT+AER 0 0 -1.183 . -1.183 . . .
## LOWCAL+AER vs NCI 0 0 -1.865 . -1.865 . . .
## LOWCAL+AER vs PHYTO/PLANT 0 0 -1.167 . -1.167 . . .
## LOWCAL+AER vs RES 0 0 -1.358 . -1.358 . . .
## LOWCAL+AER vs SC 5 1.00 -1.224 -1.224 . . . .
## LOWCAL+AER/RES vs LOWCAL+AER/RES2 1 0.65 -0.358 -1.680 2.072 -3.752 -1.66 0.0974
## LOWCAL+AER/RES vs LOWCAL+LOWFAT 0 0 -0.978 . -0.978 . . .
## LOWCAL+AER/RES vs LOWCAL+LOWFAT+AER 0 0 2.819 . 2.819 . . .
## LOWCAL+AER/RES vs LOWCAL+LOWFAT+AER2 0 0 2.096 . 2.096 . . .
## LOWCAL+AER/RES vs LOWCARB 0 0 -1.464 . -1.464 . . .
## LOWCAL+AER/RES vs LOWCARB+AER 0 0 9.669 . 9.669 . . .
## LOWCAL+AER/RES vs LOWFAT 0 0 -1.064 . -1.064 . . .
## LOWCAL+AER/RES vs LOWFAT/PHYTO+AER/RES 0 0 -0.844 . -0.844 . . .
## LOWCAL+AER/RES vs LOWFAT+AER 0 0 -0.963 . -0.963 . . .
## LOWCAL+AER/RES vs LOWFAT+PHYTO 0 0 -1.560 . -1.560 . . .
## LOWCAL+AER/RES vs MEDIT 0 0 -1.187 . -1.187 . . .
## LOWCAL+AER/RES vs MEDIT+AER 0 0 -0.890 . -0.890 . . .
## LOWCAL+AER/RES vs NCI 0 0 -1.572 . -1.572 . . .
## LOWCAL+AER/RES vs PHYTO/PLANT 0 0 -0.875 . -0.875 . . .
## LOWCAL+AER/RES vs RES 0 0 -1.065 . -1.065 . . .
## LOWCAL+AER/RES vs SC 5 1.00 -0.931 -0.931 . . . .
## LOWCAL+AER/RES2 vs LOWCAL+LOWFAT 0 0 -0.620 . -0.620 . . .
## LOWCAL+AER/RES2 vs LOWCAL+LOWFAT+AER 0 0 3.177 . 3.177 . . .
## LOWCAL+AER/RES2 vs LOWCAL+LOWFAT+AER2 0 0 2.454 . 2.454 . . .
## LOWCAL+AER/RES2 vs LOWCARB 0 0 -1.106 . -1.106 . . .
## LOWCAL+AER/RES2 vs LOWCARB+AER 0 0 10.027 . 10.027 . . .
## LOWCAL+AER/RES2 vs LOWFAT 0 0 -0.706 . -0.706 . . .
## LOWCAL+AER/RES2 vs LOWFAT/PHYTO+AER/RES 0 0 -0.486 . -0.486 . . .
## LOWCAL+AER/RES2 vs LOWFAT+AER 0 0 -0.605 . -0.605 . . .
## LOWCAL+AER/RES2 vs LOWFAT+PHYTO 0 0 -1.202 . -1.202 . . .
## LOWCAL+AER/RES2 vs MEDIT 0 0 -0.829 . -0.829 . . .
## LOWCAL+AER/RES2 vs MEDIT+AER 0 0 -0.532 . -0.532 . . .
## LOWCAL+AER/RES2 vs NCI 0 0 -1.214 . -1.214 . . .
## LOWCAL+AER/RES2 vs PHYTO/PLANT 0 0 -0.517 . -0.517 . . .
## LOWCAL+AER/RES2 vs RES 0 0 -0.707 . -0.707 . . .
## LOWCAL+AER/RES2 vs SC 1 0.88 -0.573 -1.220 4.277 -5.497 -1.66 0.0974
## LOWCAL+LOWFAT vs LOWCAL+LOWFAT+AER 0 0 3.797 . 3.797 . . .
## LOWCAL+LOWFAT vs LOWCAL+LOWFAT+AER2 0 0 3.074 . 3.074 . . .
## LOWCAL+LOWFAT vs LOWCARB 0 0 -0.486 . -0.486 . . .
## LOWCAL+LOWFAT vs LOWCARB+AER 0 0 10.647 . 10.647 . . .
## LOWCAL+LOWFAT vs LOWFAT 0 0 -0.086 . -0.086 . . .
## LOWCAL+LOWFAT vs LOWFAT/PHYTO+AER/RES 0 0 0.134 . 0.134 . . .
## LOWCAL+LOWFAT vs LOWFAT+AER 0 0 0.015 . 0.015 . . .
## LOWCAL+LOWFAT vs LOWFAT+PHYTO 0 0 -0.582 . -0.582 . . .
## LOWCAL+LOWFAT vs MEDIT 0 0 -0.209 . -0.209 . . .
## LOWCAL+LOWFAT vs MEDIT+AER 0 0 0.088 . 0.088 . . .
## LOWCAL+LOWFAT vs NCI 0 0 -0.594 . -0.594 . . .
## LOWCAL+LOWFAT vs PHYTO/PLANT 0 0 0.103 . 0.103 . . .
## LOWCAL+LOWFAT vs RES 0 0 -0.087 . -0.087 . . .
## LOWCAL+LOWFAT vs SC 1 1.00 0.047 0.047 . . . .
## LOWCAL+LOWFAT+AER vs LOWCAL+LOWFAT+AER2 1 0.89 -0.723 -0.800 -0.115 -0.685 -0.20 0.8421
## LOWCAL+LOWFAT+AER vs LOWCARB 0 0 -4.283 . -4.283 . . .
## LOWCAL+LOWFAT+AER vs LOWCARB+AER 0 0 6.850 . 6.850 . . .
## LOWCAL+LOWFAT+AER vs LOWFAT 0 0 -3.883 . -3.883 . . .
## LOWCAL+LOWFAT+AER vs LOWFAT/PHYTO+AER/RES 0 0 -3.663 . -3.663 . . .
## LOWCAL+LOWFAT+AER vs LOWFAT+AER 0 0 -3.782 . -3.782 . . .
## LOWCAL+LOWFAT+AER vs LOWFAT+PHYTO 0 0 -4.379 . -4.379 . . .
## LOWCAL+LOWFAT+AER vs MEDIT 0 0 -4.006 . -4.006 . . .
## LOWCAL+LOWFAT+AER vs MEDIT+AER 0 0 -3.709 . -3.709 . . .
## LOWCAL+LOWFAT+AER vs NCI 0 0 -4.391 . -4.391 . . .
## LOWCAL+LOWFAT+AER vs PHYTO/PLANT 0 0 -3.693 . -3.693 . . .
## LOWCAL+LOWFAT+AER vs RES 0 0 -3.884 . -3.884 . . .
## LOWCAL+LOWFAT+AER vs SC 2 1.00 -3.750 -3.750 . . . .
## LOWCAL+LOWFAT+AER2 vs LOWCARB 0 0 -3.560 . -3.560 . . .
## LOWCAL+LOWFAT+AER2 vs LOWCARB+AER 0 0 7.572 . 7.572 . . .
## LOWCAL+LOWFAT+AER2 vs LOWFAT 0 0 -3.160 . -3.160 . . .
## LOWCAL+LOWFAT+AER2 vs LOWFAT/PHYTO+AER/RES 0 0 -2.941 . -2.941 . . .
## LOWCAL+LOWFAT+AER2 vs LOWFAT+AER 0 0 -3.059 . -3.059 . . .
## LOWCAL+LOWFAT+AER2 vs LOWFAT+PHYTO 0 0 -3.656 . -3.656 . . .
## LOWCAL+LOWFAT+AER2 vs MEDIT 0 0 -3.283 . -3.283 . . .
## LOWCAL+LOWFAT+AER2 vs MEDIT+AER 0 0 -2.986 . -2.986 . . .
## LOWCAL+LOWFAT+AER2 vs NCI 0 0 -3.668 . -3.668 . . .
## LOWCAL+LOWFAT+AER2 vs PHYTO/PLANT 0 0 -2.971 . -2.971 . . .
## LOWCAL+LOWFAT+AER2 vs RES 0 0 -3.161 . -3.161 . . .
## LOWCAL+LOWFAT+AER2 vs SC 1 0.90 -3.028 -3.100 -2.388 -0.712 -0.20 0.8421
## LOWCARB vs LOWCARB+AER 0 0 11.133 . 11.133 . . .
## LOWCARB vs LOWFAT 1 1.00 0.400 0.400 . . . .
## LOWCARB vs LOWFAT/PHYTO+AER/RES 0 0 0.620 . 0.620 . . .
## LOWCARB vs LOWFAT+AER 0 0 0.501 . 0.501 . . .
## LOWCARB vs LOWFAT+PHYTO 0 0 -0.096 . -0.096 . . .
## LOWCARB vs MEDIT 0 0 0.277 . 0.277 . . .
## LOWCARB vs MEDIT+AER 0 0 0.574 . 0.574 . . .
## LOWCARB vs NCI 0 0 -0.108 . -0.108 . . .
## LOWCARB vs PHYTO/PLANT 0 0 0.590 . 0.590 . . .
## LOWCARB vs RES 0 0 0.399 . 0.399 . . .
## LOWCARB vs SC 0 0 0.533 . 0.533 . . .
## LOWCARB+AER vs LOWFAT 0 0 -10.733 . -10.733 . . .
## LOWCARB+AER vs LOWFAT/PHYTO+AER/RES 0 0 -10.513 . -10.513 . . .
## LOWCARB+AER vs LOWFAT+AER 0 0 -10.632 . -10.632 . . .
## LOWCARB+AER vs LOWFAT+PHYTO 0 0 -11.229 . -11.229 . . .
## LOWCARB+AER vs MEDIT 0 0 -10.856 . -10.856 . . .
## LOWCARB+AER vs MEDIT+AER 0 0 -10.558 . -10.558 . . .
## LOWCARB+AER vs NCI 0 0 -11.241 . -11.241 . . .
## LOWCARB+AER vs PHYTO/PLANT 0 0 -10.543 . -10.543 . . .
## LOWCARB+AER vs RES 0 0 -10.733 . -10.733 . . .
## LOWCARB+AER vs SC 1 1.00 -10.600 -10.600 . . . .
## LOWFAT vs LOWFAT/PHYTO+AER/RES 0 0 0.220 . 0.220 . . .
## LOWFAT vs LOWFAT+AER 0 0 0.101 . 0.101 . . .
## LOWFAT vs LOWFAT+PHYTO 0 0 -0.496 . -0.496 . . .
## LOWFAT vs MEDIT 1 0.03 -0.123 0.260 -0.134 0.394 0.27 0.7887
## LOWFAT vs MEDIT+AER 0 0 0.174 . 0.174 . . .
## LOWFAT vs NCI 0 0 -0.508 . -0.508 . . .
## LOWFAT vs PHYTO/PLANT 0 0 0.190 . 0.190 . . .
## LOWFAT vs RES 0 0 -0.001 . -0.001 . . .
## LOWFAT vs SC 3 1.00 0.133 0.132 1.709 -1.577 -0.75 0.4562
## LOWFAT/PHYTO+AER/RES vs LOWFAT+AER 0 0 -0.119 . -0.119 . . .
## LOWFAT/PHYTO+AER/RES vs LOWFAT+PHYTO 0 0 -0.716 . -0.716 . . .
## LOWFAT/PHYTO+AER/RES vs MEDIT 0 0 -0.343 . -0.343 . . .
## LOWFAT/PHYTO+AER/RES vs MEDIT+AER 0 0 -0.046 . -0.046 . . .
## LOWFAT/PHYTO+AER/RES vs NCI 0 0 -0.728 . -0.728 . . .
## LOWFAT/PHYTO+AER/RES vs PHYTO/PLANT 0 0 -0.030 . -0.030 . . .
## LOWFAT/PHYTO+AER/RES vs RES 0 0 -0.221 . -0.221 . . .
## LOWFAT/PHYTO+AER/RES vs SC 3 0.98 -0.087 -0.092 0.100 -0.191 -0.27 0.7860
## LOWFAT+AER vs LOWFAT+PHYTO 0 0 -0.597 . -0.597 . . .
## LOWFAT+AER vs MEDIT 0 0 -0.224 . -0.224 . . .
## LOWFAT+AER vs MEDIT+AER 0 0 0.073 . 0.073 . . .
## LOWFAT+AER vs NCI 0 0 -0.609 . -0.609 . . .
## LOWFAT+AER vs PHYTO/PLANT 0 0 0.088 . 0.088 . . .
## LOWFAT+AER vs RES 0 0 -0.102 . -0.102 . . .
## LOWFAT+AER vs SC 5 1.00 0.032 0.032 . . . .
## LOWFAT+PHYTO vs MEDIT 0 0 0.373 . 0.373 . . .
## LOWFAT+PHYTO vs MEDIT+AER 0 0 0.670 . 0.670 . . .
## LOWFAT+PHYTO vs NCI 1 1.00 -0.012 -0.011 -2.011 2.001 1.02 0.3088
## LOWFAT+PHYTO vs PHYTO/PLANT 0 0 0.685 . 0.685 . . .
## LOWFAT+PHYTO vs RES 0 0 0.495 . 0.495 . . .
## LOWFAT+PHYTO vs SC 1 0.14 0.629 -1.100 0.901 -2.001 -1.02 0.3088
## MEDIT vs MEDIT+AER 0 0 0.297 . 0.297 . . .
## MEDIT vs NCI 0 0 -0.385 . -0.385 . . .
## MEDIT vs PHYTO/PLANT 0 0 0.312 . 0.312 . . .
## MEDIT vs RES 0 0 0.122 . 0.122 . . .
## MEDIT vs SC 3 0.99 0.256 0.229 4.427 -4.198 -1.43 0.1536
## MEDIT+AER vs NCI 0 0 -0.682 . -0.682 . . .
## MEDIT+AER vs PHYTO/PLANT 0 0 0.015 . 0.015 . . .
## MEDIT+AER vs RES 0 0 -0.175 . -0.175 . . .
## MEDIT+AER vs SC 1 1.00 -0.042 -0.042 . . . .
## NCI vs PHYTO/PLANT 1 0.90 0.697 0.900 -1.101 2.001 1.02 0.3088
## NCI vs RES 0 0 0.507 . 0.507 . . .
## NCI vs SC 0 0 0.641 . 0.641 . . .
## PHYTO/PLANT vs RES 0 0 -0.190 . -0.190 . . .
## PHYTO/PLANT vs SC 1 0.97 -0.057 0.011 -1.989 2.001 1.02 0.3088
## RES vs SC 5 0.99 0.133 0.132 0.408 -0.276 -0.23 0.8201
##
## Legend:
## comparison - Treatment comparison
## k - Number of studies providing direct evidence
## prop - Direct evidence proportion
## nma - Estimated treatment effect (MD) in network meta-analysis
## direct - Estimated treatment effect (MD) derived from direct evidence
## indir. - Estimated treatment effect (MD) derived from indirect evidence
## Diff - Difference between direct and indirect treatment estimates
## z - z-value of test for disagreement (direct versus indirect)
## p-value - p-value of test for disagreement (direct versus indirect)p < 0.05 であると、非一貫である。
2.3.5 Ranking
P-スコアによるランク表を作成する。ダイエット目的のためは、低い方が良い small.values = "good。
## P-score
## NCI 0.8456
## LOWFAT+PHYTO 0.8337
## MEDIT 0.7933
## LOWFAT 0.7550
## RES 0.7503
## LOWCARB 0.7094
## LOWFAT+AER 0.6503
## LOWCAL+LOWFAT 0.6386
## SC 0.6187
## MEDIT+AER 0.6031
## PHYTO/PLANT 0.5869
## AER+RES 0.5665
## LOWFAT/PHYTO+AER/RES 0.5432
## AER 0.4664
## LOWCAL+AER/RES2 0.4543
## AER2 0.3907
## LOWCAL+AER/RES 0.2622
## LOWCAL+AER 0.2212
## LOWCAL 0.1366
## LOWCAL+LOWFAT+AER2 0.1088
## LOWCAL+LOWFAT+AER 0.0654
## LOWCARB+AER 0.0001
2.4 gemtc ベイズ
{gemtc} を使う場合には、４つの決まった列（study, mean, std.err, treatment）を持つデータフレームをあらかじめ準備する必要がある。
library(data.table)
library(gemtc)
dfGemtc <- data.table(
study = df$Study,
mean = df$Diff,
std.err = df$StdErr,
treatment = df$Treatment,
description = df$Treatment,
group = df$group)treatment では + や - など記号はダメらしい。許されているのは _ だけなので、置換する。
dfGemtc$treatment <- gsub("\\+", "_", dfGemtc$treatment)
dfGemtc$treatment <- gsub("\\-", "_", dfGemtc$treatment)
dfGemtc$treatment <- gsub("\\/", "_", dfGemtc$treatment)2.4.1 Network plot
## $Description
## [1] "MTC dataset: Network"
##
## $`Studies per treatment`
## AER AER_RES AER2
## 15 15 4
## LOWCAL LOWCAL_AER LOWCAL_AER_RES
## 2 5 5
## LOWCAL_AER_RES2 LOWCAL_LOWFAT LOWCAL_LOWFAT_AER
## 1 1 2
## LOWCAL_LOWFAT_AER2 LOWCARB LOWCARB_AER
## 1 1 1
## LOWFAT LOWFAT_AER LOWFAT_PHYTO
## 4 5 2
## LOWFAT_PHYTO_AER_RES MEDIT MEDIT_AER
## 3 3 1
## NCI PHYTO_PLANT RES
## 2 2 5
## SC
## 63
##
## $`Number of n-arm studies`
## 2-arm 3-arm
## 55 11
##
## $`Studies per treatment comparison`
## t1 t2 nr
## 1 AER AER_RES 1
## 2 AER AER2 4
## 3 AER RES 1
## 4 AER SC 15
## 5 AER_RES LOWFAT_PHYTO_AER_RES 1
## 6 AER_RES SC 15
## 7 AER2 SC 4
## 8 LOWCAL LOWFAT 1
## 9 LOWCAL SC 2
## 10 LOWCAL_AER SC 5
## 11 LOWCAL_AER_RES LOWCAL_AER_RES2 1
## 12 LOWCAL_AER_RES SC 5
## 13 LOWCAL_AER_RES2 SC 1
## 14 LOWCAL_LOWFAT SC 1
## 15 LOWCAL_LOWFAT_AER LOWCAL_LOWFAT_AER2 1
## 16 LOWCAL_LOWFAT_AER SC 2
## 17 LOWCAL_LOWFAT_AER2 SC 1
## 18 LOWCARB LOWFAT 1
## 19 LOWCARB_AER SC 1
## 20 LOWFAT MEDIT 1
## 21 LOWFAT SC 3
## 22 LOWFAT_AER SC 5
## 23 LOWFAT_PHYTO NCI 1
## 24 LOWFAT_PHYTO SC 1
## 25 LOWFAT_PHYTO_AER_RES SC 3
## 26 MEDIT SC 3
## 27 MEDIT_AER SC 1
## 28 NCI PHYTO_PLANT 1
## 29 PHYTO_PLANT SC 1
## 30 RES SC 5線の太さは調査中。
2.4.3 モデル実行
## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 143
## Unobserved stochastic nodes: 87
## Total graph size: 2050
##
## Initializing model##
## Results on the Mean Difference scale
##
## Iterations = 10:10000
## Thinning interval = 10
## Number of chains = 4
## Sample size per chain = 1000
##
## 1. Empirical mean and standard deviation for each variable,
## plus standard error of the mean:
##
## Mean SD Naive SE Time-series SE
## d.LOWFAT_PHYTO.NCI 0.01184 0.05149 0.0008141 0.0008132
## d.LOWFAT.LOWCARB 0.37575 1.57673 0.0249303 0.0249322
## d.SC.AER -0.20339 0.13889 0.0021961 0.0021694
## d.SC.AER_RES -0.06253 0.12467 0.0019712 0.0020128
## d.SC.AER2 -0.39245 0.23933 0.0037842 0.0037332
## d.SC.LOWCAL -2.39185 0.77987 0.0123308 0.0126270
## d.SC.LOWCAL_AER -1.22174 0.35079 0.0055465 0.0054476
## d.SC.LOWCAL_AER_RES -0.93057 0.28185 0.0044565 0.0043753
## d.SC.LOWCAL_AER_RES2 -0.56561 1.05430 0.0166700 0.0165211
## d.SC.LOWCAL_LOWFAT 0.03242 0.66975 0.0105896 0.0109661
## d.SC.LOWCAL_LOWFAT_AER -3.74035 0.78362 0.0123901 0.0123945
## d.SC.LOWCAL_LOWFAT_AER2 -3.03893 1.07874 0.0170564 0.0170540
## d.SC.LOWCARB_AER -10.56979 2.23603 0.0353548 0.0376567
## d.SC.LOWFAT 0.13351 0.06253 0.0009887 0.0009743
## d.SC.LOWFAT_AER 0.02739 0.18714 0.0029589 0.0029006
## d.SC.LOWFAT_PHYTO 0.61617 0.67400 0.0106569 0.0108093
## d.SC.LOWFAT_PHYTO_AER_RES -0.08821 0.10990 0.0017376 0.0016981
## d.SC.MEDIT 0.25562 0.23271 0.0036795 0.0036418
## d.SC.MEDIT_AER -0.04250 0.44750 0.0070756 0.0069895
## d.SC.PHYTO_PLANT -0.06628 0.35286 0.0055792 0.0055385
## d.SC.RES 0.13442 0.08641 0.0013663 0.0013018
##
## 2. Quantiles for each variable:
##
## 2.5% 25% 50% 75% 97.5%
## d.LOWFAT_PHYTO.NCI -0.08835 -0.02306 0.01152 0.04573 0.11536
## d.LOWFAT.LOWCARB -2.71178 -0.67066 0.37745 1.41400 3.44765
## d.SC.AER -0.48393 -0.29613 -0.19885 -0.11144 0.06150
## d.SC.AER_RES -0.30554 -0.14389 -0.06369 0.02082 0.18198
## d.SC.AER2 -0.84994 -0.55336 -0.39668 -0.23595 0.07394
## d.SC.LOWCAL -3.92108 -2.90730 -2.38965 -1.86907 -0.91452
## d.SC.LOWCAL_AER -1.90757 -1.45431 -1.21851 -0.99045 -0.55214
## d.SC.LOWCAL_AER_RES -1.47123 -1.12064 -0.92881 -0.74111 -0.37768
## d.SC.LOWCAL_AER_RES2 -2.60848 -1.28880 -0.54549 0.12962 1.51885
## d.SC.LOWCAL_LOWFAT -1.26066 -0.41796 0.02256 0.47910 1.36938
## d.SC.LOWCAL_LOWFAT_AER -5.21858 -4.26806 -3.74662 -3.21792 -2.24841
## d.SC.LOWCAL_LOWFAT_AER2 -5.15455 -3.77357 -3.02928 -2.33329 -0.93459
## d.SC.LOWCARB_AER -14.87484 -12.03606 -10.54932 -9.10729 -6.34106
## d.SC.LOWFAT 0.01823 0.09537 0.13174 0.17133 0.24615
## d.SC.LOWFAT_AER -0.32088 -0.10156 0.02950 0.15201 0.38449
## d.SC.LOWFAT_PHYTO -0.70450 0.16541 0.60412 1.06239 1.95072
## d.SC.LOWFAT_PHYTO_AER_RES -0.30876 -0.16113 -0.08794 -0.01649 0.12794
## d.SC.MEDIT -0.20691 0.10135 0.25392 0.41116 0.72290
## d.SC.MEDIT_AER -0.93457 -0.34404 -0.04964 0.25289 0.84271
## d.SC.PHYTO_PLANT -0.78254 -0.30630 -0.06296 0.16908 0.63622
## d.SC.RES -0.03939 0.07485 0.13545 0.19406 0.30074
##
## -- Model fit (residual deviance):
##
## Dbar pD DIC
## 266.72953 87.62463 354.35416
##
## 143 data points, ratio 1.865, I^2 = 47%## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 143
## Unobserved stochastic nodes: 165
## Total graph size: 2287
##
## Initializing model##
## Results on the Mean Difference scale
##
## Iterations = 5010:15000
## Thinning interval = 10
## Number of chains = 4
## Sample size per chain = 1000
##
## 1. Empirical mean and standard deviation for each variable,
## plus standard error of the mean:
##
## Mean SD Naive SE Time-series SE
## d.LOWFAT_PHYTO.NCI 0.25700 0.9371 0.014817 0.015598
## d.LOWFAT.LOWCARB 0.41990 1.8362 0.029033 0.030211
## d.SC.AER -0.37590 0.3122 0.004937 0.005029
## d.SC.AER_RES -0.28127 0.3151 0.004981 0.004821
## d.SC.AER2 -0.76719 0.5640 0.008917 0.009206
## d.SC.LOWCAL -2.91999 1.1786 0.018635 0.020043
## d.SC.LOWCAL_AER -1.53641 0.5824 0.009209 0.009107
## d.SC.LOWCAL_AER_RES -1.16902 0.5654 0.008940 0.009168
## d.SC.LOWCAL_AER_RES2 -0.56959 1.3940 0.022042 0.021852
## d.SC.LOWCAL_LOWFAT 0.01427 1.2088 0.019113 0.019118
## d.SC.LOWCAL_LOWFAT_AER -3.74887 1.0382 0.016415 0.016505
## d.SC.LOWCAL_LOWFAT_AER2 -3.05220 1.4565 0.023029 0.022762
## d.SC.LOWCARB_AER -10.57754 2.3427 0.037042 0.049127
## d.SC.LOWFAT -0.77241 0.7145 0.011297 0.011494
## d.SC.LOWFAT_AER -0.56573 0.5181 0.008192 0.008195
## d.SC.LOWFAT_PHYTO 0.04207 1.4104 0.022301 0.024682
## d.SC.LOWFAT_PHYTO_AER_RES -0.31857 0.5884 0.009303 0.009196
## d.SC.MEDIT -0.05068 0.7014 0.011091 0.011093
## d.SC.MEDIT_AER -0.04492 1.1263 0.017809 0.018228
## d.SC.PHYTO_PLANT -0.26464 0.9862 0.015593 0.015595
## d.SC.RES 0.19442 0.4941 0.007813 0.007814
## sd.d 1.00163 0.1636 0.002586 0.003300
##
## 2. Quantiles for each variable:
##
## 2.5% 25% 50% 75% 97.5%
## d.LOWFAT_PHYTO.NCI -1.5759 -0.3730 0.25625 0.88446 2.08536
## d.LOWFAT.LOWCARB -3.1927 -0.8132 0.41240 1.66227 4.04420
## d.SC.AER -0.9938 -0.5737 -0.37675 -0.17159 0.24634
## d.SC.AER_RES -0.9004 -0.4857 -0.28574 -0.07434 0.35134
## d.SC.AER2 -1.9037 -1.1359 -0.76474 -0.39578 0.36476
## d.SC.LOWCAL -5.2499 -3.7024 -2.90951 -2.11722 -0.61468
## d.SC.LOWCAL_AER -2.6831 -1.9350 -1.53928 -1.14204 -0.40561
## d.SC.LOWCAL_AER_RES -2.2790 -1.5347 -1.17020 -0.80005 -0.06168
## d.SC.LOWCAL_AER_RES2 -3.3105 -1.4873 -0.55853 0.37493 2.21778
## d.SC.LOWCAL_LOWFAT -2.4450 -0.7681 0.03449 0.81012 2.38405
## d.SC.LOWCAL_LOWFAT_AER -5.7964 -4.4380 -3.73896 -3.03847 -1.73100
## d.SC.LOWCAL_LOWFAT_AER2 -5.9241 -3.9994 -3.06288 -2.07883 -0.20416
## d.SC.LOWCARB_AER -15.1403 -12.1589 -10.58009 -9.04338 -6.10325
## d.SC.LOWFAT -2.2056 -1.2396 -0.76138 -0.28550 0.59919
## d.SC.LOWFAT_AER -1.6020 -0.8995 -0.56080 -0.21956 0.42306
## d.SC.LOWFAT_PHYTO -2.7614 -0.8891 0.05689 0.98351 2.71592
## d.SC.LOWFAT_PHYTO_AER_RES -1.4800 -0.6997 -0.31307 0.05626 0.84134
## d.SC.MEDIT -1.4973 -0.5149 -0.01339 0.41399 1.28791
## d.SC.MEDIT_AER -2.2303 -0.8049 -0.05287 0.70126 2.24581
## d.SC.PHYTO_PLANT -2.2805 -0.9138 -0.24440 0.40566 1.59774
## d.SC.RES -0.8013 -0.1207 0.19522 0.51366 1.19799
## sd.d 0.7085 0.8875 0.99144 1.10656 1.34923
##
## -- Model fit (residual deviance):
##
## Dbar pD DIC
## 149.6516 125.0159 274.6675
##
## 143 data points, ratio 1.047, I^2 = 5%
2.4.6 一貫性の評価
nodesplit <- mtc.nodesplit(mtcNetwork,
linearModel = "random",
likelihood = "normal",
link = "identity",
n.adapt = 5000,
n.iter = 1e5,
thin = 10)## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 142
## Unobserved stochastic nodes: 165
## Total graph size: 3343
##
## Initializing model
##
## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 142
## Unobserved stochastic nodes: 165
## Total graph size: 3345
##
## Initializing model
##
## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 142
## Unobserved stochastic nodes: 165
## Total graph size: 3343
##
## Initializing model
##
## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 142
## Unobserved stochastic nodes: 165
## Total graph size: 3343
##
## Initializing model
##
## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 142
## Unobserved stochastic nodes: 165
## Total graph size: 3343
##
## Initializing model
##
## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 143
## Unobserved stochastic nodes: 166
## Total graph size: 3358
##
## Initializing model
##
## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 143
## Unobserved stochastic nodes: 166
## Total graph size: 3359
##
## Initializing model
##
## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 143
## Unobserved stochastic nodes: 166
## Total graph size: 3358
##
## Initializing model
##
## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 143
## Unobserved stochastic nodes: 166
## Total graph size: 3359
##
## Initializing model
##
## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 143
## Unobserved stochastic nodes: 165
## Total graph size: 2287
##
## Initializing model
2.4.7 Ranking
LOWCARB/AER が最もよい。そう判断する根拠は、一番左の棒（１位を表す）が最も大きいため。
## SUCRA
## LOWCARB_AER 0.9995119
## LOWCAL_LOWFAT_AER 0.9134762
## LOWCAL 0.8519524
## LOWCAL_LOWFAT_AER2 0.8427381
## LOWCAL_AER 0.7226071
## LOWCAL_AER_RES 0.6486190
## AER2 0.5399048
## LOWFAT 0.5263214
## LOWFAT_AER 0.4739167
## LOWCAL_AER_RES2 0.4567738
## LOWCARB 0.4155833
## AER 0.4155714
## LOWFAT_PHYTO_AER_RES 0.3943333
## PHYTO_PLANT 0.3860357
## AER_RES 0.3778810
## MEDIT_AER 0.3348571
## LOWFAT_PHYTO 0.3289167
## LOWCAL_LOWFAT 0.3218214
## MEDIT 0.3076667
## NCI 0.2626429
## SC 0.2568452
## RES 0.2220238
2.5 BUGSnet ベイズ
2.5.2 要約
myBUGNetwork <- net.tab(data = myBUGObject,
outcome = "Diff",
N = "N",
type.outcome = "continuous",
time = NULL)
myBUGNetwork$intervention## # A tibble: 22 × 6
## Treatment n.studies n.patients min.outcome max.outcome av.outcome
## <chr> <int> <int> <dbl> <dbl> <dbl>
## 1 AER 15 729 -2 1.2 0.176
## 2 AER+RES 15 739 -3.69 3.6 0.289
## 3 AER2 4 162 -1.4 0.00866 -0.465
## 4 LOWCAL 2 27 -4 -2.7 -3.61
## 5 LOWCAL+AER 5 285 -12.2 0.301 -5.73
## 6 LOWCAL+AER/RES 5 128 -3.77 0.640 -1.93
## 7 LOWCAL+AER/RES2 1 23 -2.09 -2.09 -2.09
## 8 LOWCAL+LOWFAT 1 20 0.131 0.131 0.131
## 9 LOWCAL+LOWFAT+AER 2 78 -5.6 -4.7 -5.08
## 10 LOWCAL+LOWFAT+AER2 1 34 -4.8 -4.8 -4.8
## # … with 12 more rows2.5.3 Network plot
ラベルの順は、AER から反時計回り。最後は SC。
これは、Console で実行すると Plots に表示され、PNG や PDF で保存することができる。なお、このコードだけ実行してもうまくいかない。このファイル内の前のコードも実行する必要がある。
線が細い場合は edge.scale を大きくする。丸が小さい場合は、node.scale を大きくする。
net.plot(myBUGObject,
label.offset1 = c(5,5,5,10,5,10,5,5,5,10,5,0,5,0,5,5,5,5,5,5,5,0),
node.scale = 1,
edge.scale=2)
label.offset1 で、外側にずらすことができる。順序は、Combination から反時計回り。
2.5.5 モデル実行
set.seed(20190829)
BugsResFixed <- nma.run(BugsModelFixed,
n.adapt=1000,
n.burnin=1000,
n.iter=10000)## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 143
## Unobserved stochastic nodes: 87
## Total graph size: 1436
##
## Initializing model## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 143
## Unobserved stochastic nodes: 165
## Total graph size: 1622
##
## Initializing model2.5.6 収束の評価
## Press [ENTER] to continue plotting trace plots (or type 'stop' to end plotting)>
## Press [ENTER] to continue plotting trace plots (or type 'stop' to end plotting)>
## Press [ENTER] to continue plotting trace plots (or type 'stop' to end plotting)>
## Press [ENTER] to continue plotting trace plots (or type 'stop' to end plotting)>
## Press [ENTER] to continue plotting trace plots (or type 'stop' to end plotting)>
## Press [ENTER] to continue plotting trace plots (or type 'stop' to end plotting)>
## $gelman.rubin
## $psrf
## Point est. Upper C.I.
## d[2] 1.000855 1.003279
## d[3] 1.000131 1.000250
## d[4] 1.000146 1.000710
## d[5] 1.000297 1.001237
## d[6] 1.000327 1.001390
## d[7] 1.000206 1.000972
## d[8] 1.000147 1.000200
## d[9] 1.000541 1.002068
## d[10] 1.000584 1.002178
## d[11] 1.000034 1.000244
## d[12] 1.000179 1.000665
## d[13] 1.000463 1.001734
## d[14] 1.000005 1.000043
## d[15] 1.000671 1.002433
## d[16] 1.001382 1.004899
## d[17] 1.037845 1.095484
## d[18] 1.000219 1.000365
## d[19] 1.000487 1.001989
## d[20] 1.037620 1.094587
## d[21] 1.001993 1.006094
## d[22] 1.000106 1.000599
##
## $mpsrf
## [1] 1.018203
##
## attr(,"class")
## [1] "gelman.rubin.results"
##
## $geweke
## $stats
## Chain 1 Chain 2 Chain 3
## d[2] -1.22793886 0.84507427 0.1748677
## d[3] -0.19974493 -0.16148999 -0.4834795
## d[4] 0.15610742 0.09378271 0.7326202
## d[5] 3.37533962 -0.16656331 1.3612966
## d[6] -0.19221998 0.30668316 -0.2580536
## d[7] 0.67264434 -0.46392214 -1.0775495
## d[8] 0.18965524 -0.35625297 0.3693987
## d[9] -1.52267675 0.08757218 -0.2973917
## d[10] 0.27428249 1.51828735 -0.1837647
## d[11] -0.08355538 2.30596737 0.1927937
## d[12] -0.78334912 0.02445912 -0.5177188
## d[13] -0.56074273 1.21301400 1.1559579
## d[14] -1.63477305 -2.47980856 -0.2468580
## d[15] 0.26589271 -0.50098978 0.3023942
## d[16] -1.01085023 -1.18202421 1.6440975
## d[17] 1.62646430 1.74260110 -0.6769758
## d[18] 0.24697207 1.14101249 0.7042523
## d[19] -1.04082982 0.94172555 1.4361752
## d[20] 1.61379626 1.80413198 -0.6923763
## d[21] 4.18218071 4.28623520 -1.1254334
## d[22] -0.27936584 1.06191208 0.6665324
##
## $frac1
## [1] 0.1
##
## $frac2
## [1] 0.5
##
## attr(,"class")
## [1] "geweke.results"
## Press [ENTER] to continue plotting trace plots (or type 'stop' to end plotting)>
## Press [ENTER] to continue plotting trace plots (or type 'stop' to end plotting)>
## Press [ENTER] to continue plotting trace plots (or type 'stop' to end plotting)>
## Press [ENTER] to continue plotting trace plots (or type 'stop' to end plotting)>
## Press [ENTER] to continue plotting trace plots (or type 'stop' to end plotting)>
## Press [ENTER] to continue plotting trace plots (or type 'stop' to end plotting)>
## Press [ENTER] to continue plotting trace plots (or type 'stop' to end plotting)>
## $gelman.rubin
## $psrf
## Point est. Upper C.I.
## d[2] 1.000883 1.003166
## d[3] 1.000039 1.000081
## d[4] 1.000308 1.001221
## d[5] 1.001876 1.005977
## d[6] 1.000034 1.000295
## d[7] 1.000513 1.001914
## d[8] 1.001541 1.003873
## d[9] 1.000021 1.000295
## d[10] 1.000185 1.000340
## d[11] 1.000433 1.001476
## d[12] 1.000298 1.000351
## d[13] 1.000814 1.002194
## d[14] 1.001970 1.007012
## d[15] 1.000094 1.000258
## d[16] 1.000001 1.000156
## d[17] 1.000795 1.002073
## d[18] 1.000188 1.000682
## d[19] 1.000267 1.000825
## d[20] 1.000524 1.001148
## d[21] 1.000286 1.001210
## d[22] 1.000003 1.000096
## sigma 1.001765 1.004612
##
## $mpsrf
## [1] 1.004918
##
## attr(,"class")
## [1] "gelman.rubin.results"
##
## $geweke
## $stats
## Chain 1 Chain 2 Chain 3
## d[2] -1.1387149 -0.32109744 -1.275685835
## d[3] 0.1038653 0.06575749 0.679167788
## d[4] -1.1137324 0.33926486 0.881586871
## d[5] 0.1756987 -0.61716716 -0.850665096
## d[6] 0.6112799 0.66425924 -0.007816367
## d[7] -1.0047223 -1.20264462 -0.080682388
## d[8] -1.3196623 -0.28093493 0.034340874
## d[9] 1.0287735 -0.82265591 0.624435884
## d[10] 0.4389147 1.52071887 -0.508515900
## d[11] 0.4755243 -0.13009136 -0.470537401
## d[12] 0.1826059 -0.04684242 -1.465234264
## d[13] 0.1347647 0.37295187 0.836175124
## d[14] -0.4281452 0.04058775 -0.995622050
## d[15] 0.8228528 0.15189129 0.182046562
## d[16] -1.0368279 0.01679456 -0.153768108
## d[17] 1.0336100 -0.88980011 -0.729133309
## d[18] 1.6938124 -0.08555454 0.827789105
## d[19] -0.1939262 0.98113732 2.229989324
## d[20] 1.2592893 -0.79657844 -1.173656515
## d[21] 1.6266364 -0.32300369 -2.070883282
## d[22] -0.5506279 -1.80055611 -0.226751509
## sigma 1.0449493 0.36177835 -0.495099277
##
## $frac1
## [1] 0.1
##
## $frac2
## [1] 0.5
##
## attr(,"class")
## [1] "geweke.results"2.5.7 適合度の評価
モデルの適合度を評価する。
## $DIC
## [1] 353.0021
##
## $Dres
## [1] 266.0431
##
## $pD
## [1] 86.95902
##
## $leverage
##
## 1 0.6063229 0.5774828 0.5058759 0.2981493 0.3288017 0.380981 0.8369764
##
## 1 0.3530878 0.485577 0.5830936 0.9981901 0.9634338 0.5494841 0.60446 0.3485115
##
## 1 0.42866 0.3726717 0.870057 0.6924688 0.3563894 0.5059716 0.9144239 0.4261503
##
## 1 0.934877 0.9973525 0.673202 0.632422 0.6007998 0.7723965 0.5863764 0.5168365
##
## 1 0.5209857 0.4024556 0.6054808 0.6343793 0.5617729 0.5874121 0.5524987
##
## 1 0.3949877 0.2248529 0.5620927 0.4877972 0.7062412 0.3094156 0.9913838
##
## 1 0.7370944 0.9899069 0.925073 0.7638499 0.2208911 0.4564475 0.5807863
##
## 1 0.6540076 0.6371303 0.4693068 1.008665 0.3044174 0.4933898 0.3681409
##
## 1 0.5230902 0.8166216 0.589567 0.622958 0.5721204 0.4799247 1.006973 0.5953724
##
## 1 0.5893942 0.7074263 0.7308389 0.4961075 0.6715385 0.3107523 0.6679188
##
## 1 0.5184202 0.5889901 0.9862808 0.9626024 0.5665062 0.4712453 0.3765704
##
## 1 0.5833293 0.4212035 0.8799119 0.3668009 0.722736 0.5191738 0.1187612
##
## 1 0.2892446 0.8648639 1.003892 0.5584683 0.4731381 0.5121417 0.7459201
##
## 1 0.5903336 0.5168495 0.5962405 0.6453628 0.5887031 0.4419084 0.4822879
##
## 1 0.4509002 0.4907092 0.6400474 0.3896176 0.6025337 0.5736884 0.7002431
##
## 1 0.7291264 0.9942646 0.7752378 0.995911 0.9058777 0.7196975 0.5846969
##
## 1 0.9643337 0.4990761 0.6755949 0.7052999 0.544982 1.01116 0.4903371 0.5573537
##
## 1 0.6350837 0.4829992 0.8218932 0.5952529 0.6830653 0.5994546 0.5648885
##
## 1 1.003071 0.2087369 0.3588077 0.4422027 0.9983414 0.6549061 0.4120096 1.006574
##
## 1 0.4950603 0.2392605 0.517518 0.6143665
##
## $w
## y.1.1. y.2.1. y.3.1. y.4.1. y.5.1. y.6.1. y.7.1.
## -1.5185694 -0.8998112 -2.4641635 1.2007086 2.0356752 0.6255535 1.9684874
## y.8.1. y.9.1. y.10.1. y.11.1. y.12.1. y.13.1. y.14.1.
## -0.6430022 -0.8346272 -0.7663473 -1.0064087 -0.9891356 -0.7666993 0.9142090
## y.15.1. y.16.1. y.17.1. y.18.1. y.19.1. y.20.1. y.21.1.
## -0.6208560 -1.4987531 -0.6471575 -1.4443472 1.2402995 0.9961907 -1.0454501
## y.22.1. y.23.1. y.24.1. y.25.1. y.26.1. y.27.1. y.28.1.
## 2.0043327 2.0995421 0.9903597 0.9987105 1.6600655 0.8804338 0.9992386
## y.29.1. y.30.1. y.31.1. y.32.1. y.33.1. y.34.1. y.35.1.
## 0.8851737 1.3935510 -0.7310578 -1.7904929 -0.8732447 -0.9361461 0.8631855
## y.36.1. y.37.1. y.38.1. y.39.1. y.40.1. y.41.1. y.42.1.
## -0.7938284 1.4060940 0.9537340 -0.6555587 2.2839658 2.8653764 -0.7549912
## y.43.1. y.44.1. y.45.1. y.46.1. y.47.1. y.48.1. y.49.1.
## -1.1292098 4.7637243 0.9961052 -0.8635564 0.9949497 0.9674995 -0.9668918
## y.50.1. y.51.1. y.52.1. y.53.1. y.54.1. y.55.1. y.56.1.
## 0.5825195 1.5447196 -0.7622557 -0.8566832 -2.0679218 1.1753375 1.0043245
## y.57.1. y.58.1. y.59.1. y.60.1. y.61.1. y.62.1. y.63.1.
## -0.8073056 -0.9561221 -0.6206575 -0.7970579 0.9048670 -0.7720887 -0.8875381
## y.64.1. y.65.1. y.66.1. y.1.2. y.2.2. y.3.2. y.4.2.
## -0.8647732 -0.6928015 1.0036042 1.5470393 0.8991809 2.0063066 -1.0830079
## y.5.2. y.6.2. y.7.2. y.8.2. y.9.2. y.10.2. y.11.2.
## -1.2301703 -0.8231217 -3.5781522 0.8358592 0.8474290 0.7694108 1.0046956
## y.12.2. y.13.2. y.14.2. y.15.2. y.16.2. y.17.2. y.18.2.
## 0.9908517 0.7763135 -0.8829247 -0.6408750 1.3808477 -0.6815615 1.4629617
## y.19.2. y.20.2. y.21.2. y.22.2. y.23.2. y.24.2. y.25.2.
## -1.4533679 -1.0068393 0.8259930 -5.5631739 -1.1313092 -0.9660314 1.0019449
## y.26.2. y.27.2. y.28.2. y.29.2. y.30.2. y.31.2. y.32.2.
## -1.8386029 -0.8242105 -1.0003858 -0.8710267 -1.3980569 0.7313309 1.6813047
## y.33.2. y.34.2. y.35.2. y.36.2. y.37.2. y.38.2. y.39.2.
## 0.9270478 0.9284811 -0.7751254 0.7470944 -1.5139207 -0.9488553 0.8135922
## y.40.2. y.41.2. y.42.2. y.43.2. y.44.2. y.45.2. y.46.2.
## -1.5699266 -2.7469703 0.8072074 1.1334928 -3.1000469 -0.9973825 0.8847400
## y.47.2. y.48.2. y.49.2. y.50.2. y.51.2. y.52.2. y.53.2.
## -0.9980508 -0.9571194 0.9500470 0.9565518 0.9890999 0.7067587 0.8627482
## y.54.2. y.55.2. y.56.2. y.57.2. y.58.2. y.59.2. y.60.2.
## 1.8678589 -1.1459115 1.0055795 -0.7421597 0.9610192 0.8042575 0.7808863
## y.61.2. y.62.2. y.63.2. y.64.2. y.65.2. y.66.2. y.5.3.
## -0.9077824 0.7755284 0.9037910 0.8815667 0.7517191 -1.0015431 -1.0361175
## y.15.3. y.17.3. y.19.3. y.21.3. y.23.3. y.29.3. y.40.3.
## 0.6998336 0.7689718 -0.9991740 0.8825808 -1.4182498 -1.0032852 -0.7420236
## y.50.3. y.51.3. y.57.3.
## -1.4006907 -1.6208787 1.0273046
##
## $pmdev
## dev_a.1.1. dev_a.2.1. dev_a.3.1. dev_a.4.1. dev_a.5.1. dev_a.6.1.
## 2.3060532 0.8096602 6.0721019 1.4417012 4.1439733 0.3913172
## dev_a.7.1. dev_a.8.1. dev_a.9.1. dev_a.10.1. dev_a.11.1. dev_a.12.1.
## 3.8749425 0.4134519 0.6966026 0.5872882 1.0128584 0.9783893
## dev_a.13.1. dev_a.14.1. dev_a.15.1. dev_a.16.1. dev_a.17.1. dev_a.18.1.
## 0.5878278 0.8357781 0.3854622 2.2462608 0.4188128 2.0861388
## dev_a.19.1. dev_a.20.1. dev_a.21.1. dev_a.22.1. dev_a.23.1. dev_a.24.1.
## 1.5383428 0.9923959 1.0929658 4.0173494 4.4080771 0.9808123
## dev_a.25.1. dev_a.26.1. dev_a.27.1. dev_a.28.1. dev_a.29.1. dev_a.30.1.
## 0.9974227 2.7558173 0.7751637 0.9984778 0.7835325 1.9419844
## dev_a.31.1. dev_a.32.1. dev_a.33.1. dev_a.34.1. dev_a.35.1. dev_a.36.1.
## 0.5344455 3.2058649 0.7625563 0.8763695 0.7450892 0.6301635
## dev_a.37.1. dev_a.38.1. dev_a.39.1. dev_a.40.1. dev_a.41.1. dev_a.42.1.
## 1.9771004 0.9096085 0.4297572 5.2164998 8.2103819 0.5700118
## dev_a.43.1. dev_a.44.1. dev_a.45.1. dev_a.46.1. dev_a.47.1. dev_a.48.1.
## 1.2751148 22.6930692 0.9922256 0.7457296 0.9899249 0.9360553
## dev_a.49.1. dev_a.50.1. dev_a.51.1. dev_a.52.1. dev_a.53.1. dev_a.54.1.
## 0.9348797 0.3393290 2.3861586 0.5810338 0.7339060 4.2763007
## dev_a.55.1. dev_a.56.1. dev_a.57.1. dev_a.58.1. dev_a.59.1. dev_a.60.1.
## 1.3814181 1.0086677 0.6517424 0.9141695 0.3852157 0.6353014
## dev_a.61.1. dev_a.62.1. dev_a.63.1. dev_a.64.1. dev_a.65.1. dev_a.66.1.
## 0.8187842 0.5961209 0.7877238 0.7478327 0.4799739 1.0072215
## dev_a.1.2. dev_a.2.2. dev_a.3.2. dev_a.4.2. dev_a.5.2. dev_a.6.2.
## 2.3933305 0.8085264 4.0252662 1.1729061 1.5133190 0.6775293
## dev_a.7.2. dev_a.8.2. dev_a.9.2. dev_a.10.2. dev_a.11.2. dev_a.12.2.
## 12.8031728 0.6986606 0.7181359 0.5919930 1.0094132 0.9817870
## dev_a.13.2. dev_a.14.2. dev_a.15.2. dev_a.16.2. dev_a.17.2. dev_a.18.2.
## 0.6026627 0.7795560 0.4107208 1.9067403 0.4645261 2.1402570
## dev_a.19.2. dev_a.20.2. dev_a.21.2. dev_a.22.2. dev_a.23.2. dev_a.24.2.
## 2.1122783 1.0137254 0.6822644 30.9489033 1.2798605 0.9332166
## dev_a.25.2. dev_a.26.2. dev_a.27.2. dev_a.28.2. dev_a.29.2. dev_a.30.2.
## 1.0038936 3.3804607 0.6793230 1.0007717 0.7586875 1.9545631
## dev_a.31.2. dev_a.32.2. dev_a.33.2. dev_a.34.2. dev_a.35.2. dev_a.36.2.
## 0.5348448 2.8267855 0.8594177 0.8620771 0.6008195 0.5581501
## dev_a.37.2. dev_a.38.2. dev_a.39.2. dev_a.40.2. dev_a.41.2. dev_a.42.2.
## 2.2919560 0.9003265 0.6619323 2.4646695 7.5458459 0.6515837
## dev_a.43.2. dev_a.44.2. dev_a.45.2. dev_a.46.2. dev_a.47.2. dev_a.48.2.
## 1.2848060 9.6102906 0.9947719 0.7827648 0.9961054 0.9160776
## dev_a.49.2. dev_a.50.2. dev_a.51.2. dev_a.52.2. dev_a.53.2. dev_a.54.2.
## 0.9025893 0.9149914 0.9783186 0.4995079 0.7443345 3.4888969
## dev_a.55.2. dev_a.56.2. dev_a.57.2. dev_a.58.2. dev_a.59.2. dev_a.60.2.
## 1.3131133 1.0111901 0.5508010 0.9235579 0.6468302 0.6097834
## dev_a.61.2. dev_a.62.2. dev_a.63.2. dev_a.64.2. dev_a.65.2. dev_a.66.2.
## 0.8240688 0.6014444 0.8168382 0.7771599 0.5650816 1.0030886
## dev_a.5.3. dev_a.15.3. dev_a.17.3. dev_a.19.3. dev_a.21.3. dev_a.23.3.
## 1.0735395 0.4897670 0.5913176 0.9983487 0.7789488 2.0114325
## dev_a.29.3. dev_a.40.3. dev_a.50.3. dev_a.51.3. dev_a.57.3.
## 1.0065812 0.5505990 1.9619346 2.6272476 1.0553548
## $DIC
## [1] 274.893
##
## $Dres
## [1] 149.6142
##
## $pD
## [1] 125.2788
##
## $leverage
##
## 1 0.8988032 0.9540628 0.8376804 0.7120861 0.6550371 0.9291984 0.9567615
##
## 1 0.7849735 0.5880161 0.9564061 0.9932357 0.9924454 0.9572515 0.9068734
##
## 1 0.7880912 0.7662898 0.9010339 0.9924731 0.843809 0.6133398 0.9342777 0.997218
##
## 1 0.811939 0.9918073 1.00302 0.8957868 0.8603898 0.8134507 0.879131 0.8550764
##
## 1 0.8363575 0.9425714 0.8158897 0.9626811 0.7001798 0.9041459 0.8394724
##
## 1 0.917016 0.8000173 0.7114252 0.9731975 0.9150171 0.9611097 0.8245008 1.000376
##
## 1 0.8567207 0.9948933 0.9980646 0.9557324 0.5452404 0.7726819 0.9093028
##
## 1 0.9678342 0.9014759 0.7280166 1.010407 0.8385855 0.8993494 0.4928376
##
## 1 0.6547035 0.9393819 0.925586 0.9598195 0.9557304 0.9369643 1.004054 0.8823212
##
## 1 0.9610394 0.9039586 0.8892313 0.7831076 0.9479493 0.7776574 0.8935892
##
## 1 0.6181369 0.9607789 1.002259 0.9948885 0.9331316 0.8726117 0.7981475
##
## 1 0.8390499 0.9443625 1.01005 0.6588563 0.8272606 0.9384111 0.8666421 0.7163865
##
## 1 0.9656346 1.000158 0.8651191 0.7999123 0.7755102 0.8556048 0.8755837
##
## 1 0.8491827 0.9530067 0.8804009 0.946678 0.5540831 0.871809 0.8006002 0.9031279
##
## 1 0.8933688 0.8244476 0.9764426 0.9493705 0.9706101 0.9388507 1.005176
##
## 1 0.8934701 1.004677 1.007153 0.9530125 0.7925524 0.9499678 0.8747185 0.9596015
##
## 1 0.9208456 0.7653249 1.009062 0.911996 0.9054673 0.712921 0.6183411 0.9709663
##
## 1 0.9441354 0.9720652 0.9615042 0.9549423 0.9975798 0.5499835 0.8284853
##
## 1 0.9396294 1.013977 0.9507604 0.8172359 0.9957437 0.8550516 0.6005843
##
## 1 0.8177862 0.9334622
##
## $w
## y.1.1. y.2.1. y.3.1. y.4.1. y.5.1. y.6.1. y.7.1.
## -1.0574750 -0.9820931 -1.3520907 0.9342115 1.2188868 0.9652960 1.2145528
## y.8.1. y.9.1. y.10.1. y.11.1. y.12.1. y.13.1. y.14.1.
## -0.8942219 -0.8802325 -0.9793046 -0.9976658 -0.9980940 -0.9809916 0.9557163
## y.15.1. y.16.1. y.17.1. y.18.1. y.19.1. y.20.1. y.21.1.
## -0.8904600 -1.1044825 -0.9530440 -0.9969444 1.0367285 0.8732446 -0.9777585
## y.22.1. y.23.1. y.24.1. y.25.1. y.26.1. y.27.1. y.28.1.
## 1.1430964 1.0987114 1.0000909 -1.0015204 1.0265764 0.9325447 0.9576675
## y.29.1. y.30.1. y.31.1. y.32.1. y.33.1. y.34.1. y.35.1.
## 0.9387940 0.9798641 -0.9201621 -0.9969193 -0.9370658 -0.9846902 0.8554880
## y.36.1. y.37.1. y.38.1. y.39.1. y.40.1. y.41.1. y.42.1.
## -0.9596948 1.0091530 0.9608812 -0.9028684 1.1710694 0.9982869 -0.9565735
## y.43.1. y.44.1. y.45.1. y.46.1. y.47.1. y.48.1. y.49.1.
## -0.9959697 2.2811772 1.0004324 -0.9268364 0.9974495 0.9991184 -0.9851080
## y.50.1. y.51.1. y.52.1. y.53.1. y.54.1. y.55.1. y.56.1.
## 0.7621273 0.9650464 -0.9714487 -0.9912404 -1.1345526 0.9623622 1.0051901
## y.57.1. y.58.1. y.59.1. y.60.1. y.61.1. y.62.1. y.63.1.
## -0.9531653 -0.9683918 -0.7234253 -0.8673232 -0.9708724 -0.9777920 -0.9926623
## y.64.1. y.65.1. y.66.1. y.1.2. y.2.2. y.3.2. y.4.2.
## -0.9795805 0.9683011 1.0020275 1.0545269 0.9848318 1.2107182 -0.9774644
## y.5.2. y.6.2. y.7.2. y.8.2. y.9.2. y.10.2. y.11.2.
## -0.9147932 -0.9738294 -1.7155528 0.9508564 0.8923379 0.9813344 1.0021236
## y.12.2. y.13.2. y.14.2. y.15.2. y.16.2. y.17.2. y.18.2.
## 0.9996426 0.9688099 -0.9375423 0.8936181 1.0804335 -0.9717868 1.0058397
## y.19.2. y.20.2. y.21.2. y.22.2. y.23.2. y.24.2. y.25.2.
## -1.0676509 -0.9467886 0.9693863 -2.0044145 -0.9404861 -0.9904277 1.0000996
## y.26.2. y.27.2. y.28.2. y.29.2. y.30.2. y.31.2. y.32.2.
## -1.0395262 -0.9005757 -0.9455687 -0.9268916 -0.9868545 0.9288655 1.0005428
## y.33.2. y.34.2. y.35.2. y.36.2. y.37.2. y.38.2. y.39.2.
## 0.9587809 0.9768295 -0.7767195 0.9439676 -1.0166312 -0.9523155 0.9501106
## y.40.2. y.41.2. y.42.2. y.43.2. y.44.2. y.45.2. y.46.2.
## -1.0569479 -0.9986755 0.9744515 1.0024055 -1.6422439 -1.0026684 0.9462464
## y.47.2. y.48.2. y.49.2. y.50.2. y.51.2. y.52.2. y.53.2.
## -1.0023358 -1.0035826 0.9843445 0.9377274 0.9771178 0.9541465 0.9864877
## y.54.2. y.55.2. y.56.2. y.57.2. y.58.2. y.59.2. y.60.2.
## 1.1036176 -0.9680194 -1.0045723 -0.9556674 0.9706572 0.8541858 0.8544357
## y.61.2. y.62.2. y.63.2. y.64.2. y.65.2. y.66.2. y.5.3.
## 0.9870342 0.9852642 0.9976835 0.9827607 -0.9773422 -0.9988065 -1.0937902
## y.15.3. y.17.3. y.19.3. y.21.3. y.23.3. y.29.3. y.40.3.
## 0.9113973 0.9744482 1.0069915 0.9831328 -0.9558523 0.9978697 -0.9292904
## y.50.3. y.51.3. y.57.3.
## -1.0165085 -1.0111071 0.9899559
##
## $pmdev
## dev_a.1.1. dev_a.2.1. dev_a.3.1. dev_a.4.1. dev_a.5.1. dev_a.6.1.
## 1.1182533 0.9645068 1.8281492 0.8727511 1.4856850 0.9317964
## dev_a.7.1. dev_a.8.1. dev_a.9.1. dev_a.10.1. dev_a.11.1. dev_a.12.1.
## 1.4751385 0.7996328 0.7748092 0.9590375 0.9953371 0.9961916
## dev_a.13.1. dev_a.14.1. dev_a.15.1. dev_a.16.1. dev_a.17.1. dev_a.18.1.
## 0.9623446 0.9133937 0.7929191 1.2198815 0.9082929 0.9938982
## dev_a.19.1. dev_a.20.1. dev_a.21.1. dev_a.22.1. dev_a.23.1. dev_a.24.1.
## 1.0748060 0.7625562 0.9560118 1.3066695 1.2071667 1.0001818
## dev_a.25.1. dev_a.26.1. dev_a.27.1. dev_a.28.1. dev_a.29.1. dev_a.30.1.
## 1.0030432 1.0538591 0.8696396 0.9171271 0.8813342 0.9601337
## dev_a.31.1. dev_a.32.1. dev_a.33.1. dev_a.34.1. dev_a.35.1. dev_a.36.1.
## 0.8466983 0.9938481 0.8780923 0.9696149 0.7318598 0.9210141
## dev_a.37.1. dev_a.38.1. dev_a.39.1. dev_a.40.1. dev_a.41.1. dev_a.42.1.
## 1.0183897 0.9232927 0.8151713 1.3714036 0.9965766 0.9150329
## dev_a.43.1. dev_a.44.1. dev_a.45.1. dev_a.46.1. dev_a.47.1. dev_a.48.1.
## 0.9919556 5.2037694 1.0008650 0.8590257 0.9949054 0.9982375
## dev_a.49.1. dev_a.50.1. dev_a.51.1. dev_a.52.1. dev_a.53.1. dev_a.54.1.
## 0.9704377 0.5808380 0.9313146 0.9437126 0.9825574 1.2872097
## dev_a.55.1. dev_a.56.1. dev_a.57.1. dev_a.58.1. dev_a.59.1. dev_a.60.1.
## 0.9261410 1.0104072 0.9085241 0.9377828 0.5233442 0.7522496
## dev_a.61.1. dev_a.62.1. dev_a.63.1. dev_a.64.1. dev_a.65.1. dev_a.66.1.
## 0.9425932 0.9560772 0.9853784 0.9595780 0.9376071 1.0040591
## dev_a.1.2. dev_a.2.2. dev_a.3.2. dev_a.4.2. dev_a.5.2. dev_a.6.2.
## 1.1120269 0.9698937 1.4658387 0.9554366 0.8368466 0.9483437
## dev_a.7.2. dev_a.8.2. dev_a.9.2. dev_a.10.2. dev_a.11.2. dev_a.12.2.
## 2.9431215 0.9041278 0.7962669 0.9630172 1.0042517 0.9992853
## dev_a.13.2. dev_a.14.2. dev_a.15.2. dev_a.16.2. dev_a.17.2. dev_a.18.2.
## 0.9385926 0.8789855 0.7985533 1.1673366 0.9443696 1.0117135
## dev_a.19.2. dev_a.20.2. dev_a.21.2. dev_a.22.2. dev_a.23.2. dev_a.24.2.
## 1.1398785 0.8964087 0.9397098 4.0176776 0.8845140 0.9809471
## dev_a.25.2. dev_a.26.2. dev_a.27.2. dev_a.28.2. dev_a.29.2. dev_a.30.2.
## 1.0001992 1.0806146 0.8110366 0.8941002 0.8591280 0.9738817
## dev_a.31.2. dev_a.32.2. dev_a.33.2. dev_a.34.2. dev_a.35.2. dev_a.36.2.
## 0.8627912 1.0010859 0.9192608 0.9541958 0.6032932 0.8910748
## dev_a.37.2. dev_a.38.2. dev_a.39.2. dev_a.40.2. dev_a.41.2. dev_a.42.2.
## 1.0335390 0.9069047 0.9027102 1.1171389 0.9973528 0.9495557
## dev_a.43.2. dev_a.44.2. dev_a.45.2. dev_a.46.2. dev_a.47.2. dev_a.48.2.
## 1.0048168 2.6969651 1.0053440 0.8953822 1.0046770 1.0071780
## dev_a.49.2. dev_a.50.2. dev_a.51.2. dev_a.52.2. dev_a.53.2. dev_a.54.2.
## 0.9689340 0.8793327 0.9547592 0.9103956 0.9731580 1.2179718
## dev_a.55.2. dev_a.56.2. dev_a.57.2. dev_a.58.2. dev_a.59.2. dev_a.60.2.
## 0.9370616 1.0091654 0.9133002 0.9421754 0.7296334 0.7300604
## dev_a.61.2. dev_a.62.2. dev_a.63.2. dev_a.64.2. dev_a.65.2. dev_a.66.2.
## 0.9742365 0.9707456 0.9953724 0.9658186 0.9551978 0.9976144
## dev_a.5.3. dev_a.15.3. dev_a.17.3. dev_a.19.3. dev_a.21.3. dev_a.23.3.
## 1.1963769 0.8306450 0.9495493 1.0140318 0.9665501 0.9136536
## dev_a.29.3. dev_a.40.3. dev_a.50.3. dev_a.51.3. dev_a.57.3.
## 0.9957440 0.8635807 1.0332895 1.0223376 0.9800126DIC 値が低いほど、適合度が高い。ランダム効果の方が良い。
References
LeVasseur, Nathalie, Wei Cheng, Sasha Mazzarello, Mark Clemons, Lisa Vandermeer, Lee Jones, Anil Abraham Joy, et al. 2021. “Optimising Weight-Loss Interventions in Cancer Patients—a Systematic Review and Network Meta-Analysis.” PloS One 16 (2): e0245794.