| CN0100013261 |
52.00 |
0.50 |
@@ -4940,7 +4903,7 @@ Clustering
),
pvd, y_shift = -3.5
)
-
+
wss <- sapply(
list(
nn.min = agcs$cluster.nn.min,
@@ -4958,7 +4921,7 @@ Clustering
main = "Within cluster sum of squares"
)
legend("topright", legend = colnames(wss), col = col, pch = 15)
-
+
barplot(
1 - colSums(wss) / tss, col = col, horiz = TRUE,
main = "Between/Total sum of squares"
@@ -4969,7 +4932,7 @@ Clustering
col = col, pch = 15,
bty = "n"
)
-
+
Lineare Regression
@@ -4978,8 +4941,8 @@
Lineare Regression
agcs$cluster.hc.wardD2 %in% which(table(agcs$cluster.hc.wardD2) == clsize)
table(keep)
## keep
-## TRUE
-## 120
+## FALSE TRUE
+## 19 100
##
## Call:
## lm(formula = weight ~ 0 + n.cases.lama.nonlap.tiva + n.cases.tube.nonlap.tiva +
@@ -4992,34 +4955,34 @@ Lineare Regression
##
## Residuals:
## Min 1Q Median 3Q Max
-## -187.74 -32.27 17.88 58.42 331.33
+## -139.74 -40.13 13.32 55.49 332.02
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
-## n.cases.lama.nonlap.tiva 6.462e-01 3.636e+00 0.178 0.859311
-## n.cases.tube.nonlap.tiva -8.740e+00 4.696e+00 -1.861 0.065608 .
-## n.cases.tube.lap.tiva -8.973e+00 2.424e+01 -0.370 0.711984
-## n.cases.lama.nonlap.inha -1.770e+00 2.772e+00 -0.639 0.524447
-## n.cases.tube.nonlap.inha 6.983e+00 1.568e+00 4.454 2.16e-05 ***
-## n.cases.tube.lap.inha 8.201e-01 8.081e-01 1.015 0.312582
-## sum.dura.lama.nonlap.tiva -1.068e-02 6.042e-02 -0.177 0.859997
-## sum.dura.tube.nonlap.tiva 9.534e-02 3.990e-02 2.390 0.018697 *
-## sum.dura.tube.lap.tiva -1.928e-01 2.483e-01 -0.777 0.439191
-## sum.dura.lama.nonlap.inha 6.282e-02 4.318e-02 1.455 0.148831
-## sum.dura.tube.nonlap.inha -4.309e-02 1.176e-02 -3.664 0.000397 ***
-## sum.dura.tube.lap.inha -2.398e-04 1.968e-04 -1.218 0.225965
-## med.flow.lama.nonlap.tiva -4.417e+00 6.615e+00 -0.668 0.505874
-## med.flow.tube.nonlap.tiva 6.477e+01 1.998e+01 3.241 0.001607 **
-## med.flow.tube.lap.tiva 4.940e+01 3.310e+01 1.492 0.138741
-## med.flow.lama.nonlap.inha 6.936e+01 2.209e+01 3.139 0.002214 **
-## med.flow.tube.nonlap.inha 4.055e+02 4.719e+01 8.593 1.04e-13 ***
-## med.flow.tube.lap.inha 1.327e+02 4.964e+01 2.673 0.008760 **
+## n.cases.lama.nonlap.tiva 3.961521 5.175262 0.765 0.44619
+## n.cases.tube.nonlap.tiva -9.306865 9.127342 -1.020 0.31088
+## n.cases.tube.lap.tiva -3.476877 34.769426 -0.100 0.92059
+## n.cases.lama.nonlap.inha -1.569260 3.401434 -0.461 0.64577
+## n.cases.tube.nonlap.inha 6.267255 2.879208 2.177 0.03238 *
+## n.cases.tube.lap.inha -1.313977 2.142855 -0.613 0.54145
+## sum.dura.lama.nonlap.tiva -0.073381 0.089459 -0.820 0.41443
+## sum.dura.tube.nonlap.tiva 0.046780 0.070891 0.660 0.51117
+## sum.dura.tube.lap.tiva -0.196741 0.261873 -0.751 0.45463
+## sum.dura.lama.nonlap.inha 0.054540 0.053887 1.012 0.31446
+## sum.dura.tube.nonlap.inha -0.026262 0.026016 -1.009 0.31571
+## sum.dura.tube.lap.inha 0.004923 0.009953 0.495 0.62217
+## med.flow.lama.nonlap.tiva -5.608559 7.410623 -0.757 0.45132
+## med.flow.tube.nonlap.tiva 70.319102 25.197113 2.791 0.00654 **
+## med.flow.tube.lap.tiva 36.352732 36.216645 1.004 0.31845
+## med.flow.lama.nonlap.inha 75.588341 26.670177 2.834 0.00578 **
+## med.flow.tube.nonlap.inha 421.095724 55.683562 7.562 5.11e-11 ***
+## med.flow.tube.lap.inha 144.071343 57.421952 2.509 0.01408 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
-## Residual standard error: 100.2 on 102 degrees of freedom
-## Multiple R-squared: 0.9466, Adjusted R-squared: 0.9372
-## F-statistic: 100.5 on 18 and 102 DF, p-value: < 2.2e-16
+## Residual standard error: 103.3 on 82 degrees of freedom
+## Multiple R-squared: 0.9453, Adjusted R-squared: 0.9333
+## F-statistic: 78.76 on 18 and 82 DF, p-value: < 2.2e-16
Penalized Regression
@@ -5128,34 +5091,34 @@
Penalized Regression
))
out
## $sd_infl
-## [1] 1.582413
+## [1] 1
##
## $err_hat
-## [1] 15354
+## [1] 14144.56
##
## $ci_lo
-## [1] 6734.32
+## [1] 8280.237
##
## $ci_hi
-## [1] 23973.69
+## [1] 20008.88
##
## $raw_mean
-## [1] 16707.71
+## [1] 16408.87
##
## $bias_est
-## [1] 1353.708
+## [1] 2264.31
##
## $sd
-## [1] 27230.72
+## [1] 26761.75
##
## $running_sd_infl
-## [1] 0.0000000 0.9894073 1.1455468 1.2183734 1.5824130
+## [1] 1.9703676 1.3969710 0.9897224 0.9800462 0.8761885
##
## $ho_err
-## [1] 18600.89
+## [1] 17481.92
# RMSE
sqrt(c(out$err_hat, out$ho_err))
-
## [1] 123.9113 136.3851
+
## [1] 118.9309 132.2192
knitr::kable(
cbind(
as.matrix(coef(fit, s = "lambda.1se")),
@@ -5195,17 +5158,17 @@ Penalized Regression
| n.cases.lama.nonlap.inha |
0.0000000 |
-0.5253547 |
+0.1965923 |
| n.cases.tube.nonlap.inha |
-1.2574932 |
-2.4844146 |
+0.9972459 |
+2.3741882 |
| n.cases.tube.lap.inha |
0.0000000 |
-0.7505375 |
+0.0000000 |
| sum.dura.lama.nonlap.tiva |
@@ -5215,21 +5178,21 @@ Penalized Regression
| sum.dura.tube.nonlap.tiva |
0.0000000 |
-0.0118915 |
+0.0000000 |
| sum.dura.tube.lap.tiva |
0.0000000 |
--0.0566185 |
+0.0000000 |
| sum.dura.lama.nonlap.inha |
0.0000000 |
-0.0133827 |
+0.0000000 |
| sum.dura.tube.nonlap.inha |
-0.0021453 |
+0.0011769 |
0.0000000 |
@@ -5245,7 +5208,7 @@ Penalized Regression
| med.flow.tube.nonlap.tiva |
0.0000000 |
-6.4202110 |
+21.9898359 |
| med.flow.tube.lap.tiva |
@@ -5255,24 +5218,24 @@ Penalized Regression
| med.flow.lama.nonlap.inha |
0.0000000 |
-52.2273947 |
+24.5337392 |
| med.flow.tube.nonlap.inha |
-625.0801441 |
-518.5774055 |
+633.2554403 |
+580.0182195 |
| med.flow.tube.lap.inha |
0.0000000 |
-58.6741698 |
+0.0000000 |
