1 tabla univariante

Data Frame Summary

datisc

Dimensions: 63 x 48
Duplicates: 0
No Variable Stats / Values Freqs (% of Valid) Graph Valid Missing
1 articulacion [factor]
1. CADERA
2. RODILLA
32(50.8%)
31(49.2%)
 63 (100.0%) 0 (0.0%)
2 tipo_intervencion [factor]
1. PRIMARIA
2. REVISIÓN
51(81.0%)
12(19.0%)
 63 (100.0%) 0 (0.0%)
3 indicacion [factor]
1. Artrosis
2. Other
43(68.3%)
20(31.7%)
 63 (100.0%) 0 (0.0%)
4 cementada [factor]
1. Cementado
2. NO
34(54.0%)
29(46.0%)
 63 (100.0%) 0 (0.0%)
5 polietileno [factor]
1. NO
2. SI
30(47.6%)
33(52.4%)
 63 (100.0%) 0 (0.0%)
6 edad [numeric]
Mean (sd) : 66.3 (12.3)
min ≤ med ≤ max:
16 ≤ 68 ≤ 85
IQR (CV) : 18 (0.2)
 35 distinct values 63 (100.0%) 0 (0.0%)
7 sexo [factor]
1. H
2. M
43(68.3%)
20(31.7%)
 63 (100.0%) 0 (0.0%)
8 asa [factor]
1. I
2. II
3. III
4. IV
7(11.1%)
34(54.0%)
21(33.3%)
1(1.6%)
 63 (100.0%) 0 (0.0%)
9 ecv [factor]
1. FALSE
2. TRUE
42(66.7%)
21(33.3%)
 63 (100.0%) 0 (0.0%)
10 ete_previa [factor]
1. FALSE
2. TRUE
58(92.1%)
5(7.9%)
 63 (100.0%) 0 (0.0%)
11 neumopatia [factor]
1. FALSE
2. TRUE
48(76.2%)
15(23.8%)
 63 (100.0%) 0 (0.0%)
12 hta [factor]
1. FALSE
2. TRUE
25(39.7%)
38(60.3%)
 63 (100.0%) 0 (0.0%)
13 psiquiatrica [factor]
1. FALSE
2. TRUE
61(96.8%)
2(3.2%)
 63 (100.0%) 0 (0.0%)
14 diabetes [factor]
1. FALSE
2. TRUE
45(71.4%)
18(28.6%)
 63 (100.0%) 0 (0.0%)
15 imc [numeric]
Mean (sd) : 31.7 (6)
min ≤ med ≤ max:
19 ≤ 32 ≤ 51
IQR (CV) : 7.5 (0.2)
 24 distinct values 63 (100.0%) 0 (0.0%)
16 gastropatia [factor]
1. FALSE
2. TRUE
52(82.5%)
11(17.5%)
 63 (100.0%) 0 (0.0%)
17 inmunodepresion [factor]
1. FALSE
2. TRUE
59(93.7%)
4(6.3%)
 63 (100.0%) 0 (0.0%)
18 malignidad [factor]
1. F
2. V
53(84.1%)
10(15.9%)
 63 (100.0%) 0 (0.0%)
19 marcapasos [factor]
1. F
2. V
62(98.4%)
1(1.6%)
 63 (100.0%) 0 (0.0%)
20 renopatia [factor]
1. F
2. V
56(88.9%)
7(11.1%)
 63 (100.0%) 0 (0.0%)
21 reumatismo [factor]
1. F
2. V
51(81.0%)
12(19.0%)
 63 (100.0%) 0 (0.0%)
22 hepatopatia [factor]
1. F
2. V
61(96.8%)
2(3.2%)
 63 (100.0%) 0 (0.0%)
23 iproprevia [factor]
1. F
2. V
62(98.4%)
1(1.6%)
 63 (100.0%) 0 (0.0%)
24 vih [factor] 1. F
63(100.0%)
 63 (100.0%) 0 (0.0%)
25 coagulopatia [factor]
1. F
2. V
54(85.7%)
9(14.3%)
 63 (100.0%) 0 (0.0%)
26 anticoagulantes [factor]
1. F
2. V
56(88.9%)
7(11.1%)
 63 (100.0%) 0 (0.0%)
27 antiagregacion [factor]
1. FALSE
2. TRUE
46(75.4%)
15(24.6%)
 61 (96.8%) 2 (3.2%)
28 tabaco [factor]
1. FALSE
2. TRUE
3. VERDADERO
4. VERDADERO EX
33(52.4%)
11(17.5%)
1(1.6%)
18(28.6%)
 63 (100.0%) 0 (0.0%)
29 alcohol [factor]
1. FALSE
2. TRUE
50(79.4%)
13(20.6%)
 63 (100.0%) 0 (0.0%)
30 seroma [factor]
1. FALSE
2. TRUE
27(42.9%)
36(57.1%)
 63 (100.0%) 0 (0.0%)
31 hematoma [factor]
1. FALSE
2. TRUE
45(71.4%)
18(28.6%)
 63 (100.0%) 0 (0.0%)
32 skin_infection [factor]
1. FALSE
2. TRUE
34(54.0%)
29(46.0%)
 63 (100.0%) 0 (0.0%)
33 fistula [factor]
1. FALSE
2. TRUE
40(63.5%)
23(36.5%)
 63 (100.0%) 0 (0.0%)
34 fiebre [factor]
1. FALSE
2. TRUE
41(65.1%)
22(34.9%)
 63 (100.0%) 0 (0.0%)
35 tipo_ip [factor]
1. AGUDA
2. AGUDA > CRÓNICA
3. HEMATÓGENA
53(84.1%)
1(1.6%)
9(14.3%)
 63 (100.0%) 0 (0.0%)
36 diasadair [numeric]
Mean (sd) : 35.3 (17.7)
min ≤ med ≤ max:
13 ≤ 29.5 ≤ 87
IQR (CV) : 20.8 (0.5)
 34 distinct values 54 (85.7%) 9 (14.3%)
37 dias_clinica [numeric]
Mean (sd) : 15.3 (13.7)
min ≤ med ≤ max:
2 ≤ 8 ≤ 43
IQR (CV) : 25 (0.9)
 15 distinct values 21 (33.3%) 42 (66.7%)
38 pcr_sangre [numeric]
Mean (sd) : 13.5 (23.6)
min ≤ med ≤ max:
0.1 ≤ 7.9 ≤ 180
IQR (CV) : 13.1 (1.7)
 56 distinct values 63 (100.0%) 0 (0.0%)
39 wbc_sangre [numeric]
Mean (sd) : 9.3 (4.1)
min ≤ med ≤ max:
3.2 ≤ 8.2 ≤ 24
IQR (CV) : 4.6 (0.4)
 47 distinct values 63 (100.0%) 0 (0.0%)
40 hemocultivo [factor]
1. FALSE
2. TRUE
59(93.7%)
4(6.3%)
 63 (100.0%) 0 (0.0%)
41 germen [factor]
1. 0
2. CULTIVO NEGATIVO
3. Enterococo + Serratia
4. Enterococus f
5. Listeria m
6. MRSA
7. Propionibacterium acidifa
8. Pseudomona
9. S capitis + OTRO SCN NO I
10. S.capitis
[ 24 others ]
5(8.2%)
1(1.6%)
1(1.6%)
1(1.6%)
1(1.6%)
2(3.3%)
1(1.6%)
1(1.6%)
1(1.6%)
1(1.6%)
46(75.4%)
 61 (96.8%) 2 (3.2%)
42 cultivoporc [numeric]
Mean (sd) : 62.4 (34.9)
min ≤ med ≤ max:
0 ≤ 66.7 ≤ 100
IQR (CV) : 66.7 (0.6)
 18 distinct values 63 (100.0%) 0 (0.0%)
43 resdair [factor]
1. FALSE
2. TRUE
25(39.7%)
38(60.3%)
 63 (100.0%) 0 (0.0%)
44 klic_score [numeric]
Mean (sd) : 2.1 (1.8)
min ≤ med ≤ max:
0 ≤ 2 ≤ 6.5
IQR (CV) : 3.5 (0.9)
 10 distinct values 41 (65.1%) 22 (34.9%)
45 shohat [numeric]
Mean (sd) : 64.6 (9.4)
min ≤ med ≤ max:
44.2 ≤ 65.6 ≤ 86.9
IQR (CV) : 15.6 (0.1)
 52 distinct values 54 (85.7%) 9 (14.3%)
46 edadmas70 [logical]
1. FALSE
2. TRUE
39(61.9%)
24(38.1%)
 63 (100.0%) 0 (0.0%)
47 imc30 [logical]
1. FALSE
2. TRUE
23(36.5%)
40(63.5%)
 63 (100.0%) 0 (0.0%)
48 asamas2 [logical]
1. FALSE
2. TRUE
41(65.1%)
22(34.9%)
 63 (100.0%) 0 (0.0%)

Generated by summarytools 1.0.0 (R version 4.0.2)
2021-11-29

2 grafico continuas

## Warning: Removed 7 rows containing non-finite values (stat_density).
## Warning: Removed 1 rows containing non-finite values (stat_density).
## Warning: Removed 9 rows containing non-finite values (stat_density).
## Warning: Removed 42 rows containing non-finite values (stat_density).

3 tabla bivariante

4 validaciones del klic_score y shothat

4.1 muestra y parametros previos

Se dispone de una muestra de 64 personas para shothat, 41 para klic_score. El tamaño necesario para una validación es un asunto controvertido. Harrell et al sugieren que al menos haya 100 casos (). Vergouwe et al. que, para un resultado binario, al menos, a estos 100 casos, se acompañe de 100 controles (). Estas recomendaciones se basaban en el tamaño muestral necesario para detectar una diferencia significativa entre las medidas de rendimiento detectadas y las pre-especificadas con un poder del 80% y 5% de nivel de significación (por ejemplo, asumiendo una diferencia de 0.1 en el estadistico C), y asumiendo que la prevalencia se mantenia constante.. Estudios recientes han abordado una relajacion de estas asunciones mediante simulacion (Riley, Pavlou)

Harrell FE, Lee KL and Mark DB. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 1996; 15: 361–387.

Vergouwe Y, Steyerberg EW, Eijkemans MJC, et al. Substantial effective sample sizes were required for external validation studies of predictive logistic regression models. J Clin Epidemiol 2005; 58: 475–483.

Pavlou M, Qu C, Omar RZ, Seaman SR, Steyerberg EW, White IR, Ambler G. Estimation of required sample size for external validation of risk models for binary outcomes. Statistical Methods in Medical Research. 2021 Apr

Riley RD, Debray TP, Collins GS, Archer L, Ensor J, van Smeden M, Snell KI. Minimum sample size for external validation of a clinical prediction model with a binary outcome. Statistics in Medicine. 2021 May 24.

## siguiendo el trabajo de Pavlou:

#The target is to calculate the size of the validation data so as to estimate the C-statistic, the Calibration Slope and the Calibration in the Large with sufficient precision. In this example, the required precision is reflected by a SE of the estimated C-statistic of at most 0.025, and SE of the estimated Calibration Slope and Calibration in the Large of at most 0.1.

# The anticipated values of the outcome prevalence and the C-statistic are p=0. and C=0.75, respectively.

sampsizeval(p=0.4, c=0.76, se_c=0.06, se_cs =0.2, se_cl = 0.15)
## $size_c_statistic
## [1] 65
## 
## $size_calibration_slope
## [1] 159
## 
## $size_calibration_large
## [1] 231
## 
## $size_recommended
## [1] 231

Se concluye que con 64 personas asumiendo una prevalencia (incidencia de fracaso) de 0.4, podriamos detectar un estadistico C de 0.76 con una precision de 0.06.

4.2 shohat

4.2.1 Performance Measure

perfMeasures(preds, truth = as.numeric(datiscs$resdair), namePos = 1)
## 
##      Performance Measure(s)
## 
##                                        Measure       Value
## 1                               accuracy (ACC)  0.31481481
## 2  probability of correct classification (PCC)  0.31481481
## 3                        fraction correct (FC)  0.31481481
## 4            simple matching coefficient (SMC)  0.31481481
## 5                Rand (similarity) index (RSI)  0.31481481
## 6       probability of misclassification (PMC)  0.68518519
## 7                              error rate (ER)  0.68518519
## 8                     fraction incorrect (FIC)  0.68518519
## 9                           sensitivity (SENS)  0.52380952
## 10                                recall (REC)  0.52380952
## 11                    true positive rate (TPR)  0.52380952
## 12               probability of detection (PD)  0.52380952
## 13                               hit rate (HR)  0.52380952
## 14                          specificity (SPEC)  0.18181818
## 15                    true negative rate (TNR)  0.18181818
## 16                           selectivity (SEL)  0.18181818
## 17                         detection rate (DR)  0.20370370
## 18                   false positive rate (FPR)  0.81818182
## 19                               fall-out (FO)  0.81818182
## 20                    false alarm (rate) (FAR)  0.81818182
## 21            probability of false alarm (PFA)  0.81818182
## 22                   false negative rate (FNR)  0.47619048
## 23                              miss rate (MR)  0.47619048
## 24                  false discovery rate (FDR)  0.71052632
## 25                   false omission rate (FOR)  0.62500000
## 26                           prevalence (PREV)  0.38888889
## 27      (positive) pre-test probability (PREP)  0.38888889
## 28             (positive) pre-test odds (PREO)  0.63636364
## 29                detection prevalence (DPREV)  0.70370370
## 30       negative pre-test probability (NPREP)  0.61111111
## 31              negative pre-test odds (NPREO)  1.57142857
## 32                   no information rate (NIR)  0.61111111
## 33                    weighted accuracy (WACC)  0.35281385
## 34                    balanced accuracy (BACC)  0.35281385
## 35              (bookmaker) informedness (INF) -0.29437229
## 36                  Youden's J statistic (YJS) -0.29437229
## 37                               deltap' (DPp) -0.29437229
## 38             positive likelihood ratio (PLR)  0.64021164
## 39             negative likelihood ratio (NLR)  2.61904762
## 40             weighted likelihood ratio (WLR)  1.62962963
## 41             balanced likelihood ratio (BLR)  1.62962963
## 42                 diagnostic odds ratio (DOR)  0.24444444
## 43             positive predictive value (PPV)  0.28947368
## 44                            precision (PREC)  0.28947368
## 45    (positive) post-test probability (POSTP)  0.28947368
## 46           (positive) post-test odds (POSTO)  0.40740741
## 47                      Bayes factor G1 (BFG1)  0.64021164
## 48             negative predictive value (NPV)  0.37500000
## 49     negative post-test probability (NPOSTP)  0.37500000
## 50            negative post-test odds (NPOSTO)  0.60000000
## 51                      Bayes factor G0 (BFG0)  0.38181818
## 52                           markedness (MARK) -0.33552632
## 53                                 deltap (DP) -0.33552632
## 54             weighted predictive value (WPV)  0.33223684
## 55             balanced predictive value (BPV)  0.33223684
## 56                              F1 score (F1S)  0.37288136
## 57           Dice similarity coefficient (DSC)  0.37288136
## 58                          F beta score (FBS)  0.37288136
## 59        Jaccard similarity coefficient (JSC)  0.22916667
## 60                           threat score (TS)  0.22916667
## 61                critical success index (CSI)  0.22916667
## 62     Matthews' correlation coefficient (MCC) -0.31427639
## 63        Pearson's correlation (r phi) (RPHI) -0.31427639
## 64                      Phi coefficient (PHIC) -0.31427639
## 65                            Cramer's V (CRV) -0.31427639
## 66    proportion of positive predictions (PPP)  0.70370370
## 67                    expected accuracy (EACC)  0.45473251
## 68             Cohen's kappa coefficient (CKC) -0.25660377
## 69            mutual information in bits (MI2)  0.07044026
## 70                 joint entropy in bits (JE2)  1.22711643
## 71      variation of information in bits (VI2)  1.69991453
## 72                       Jaccard distance (JD)  0.96021122
## 73           information quality ratio (INFQR)  0.03978878
## 74                uncertainty coefficient (UC)  0.07306484
## 75                    entropy coefficient (EC)  0.07306484
## 76                 proficiency (metric) (PROF)  0.07306484
## 77                   deficiency (metric) (DFM)  0.92693516
## 78                            redundancy (RED)  0.03826622
## 79                  symmetric uncertainty (SU)  0.07653243
## 80                 normalized uncertainty (NU)  0.07661877

4.2.2 Performance Score and optimal cutoff

perfScores(preds, truth = as.numeric(datiscs$resdair), namePos = 1)
## Warning in AUC(pred, group = as.integer(truth == namePos)): The computed
## AUC value 0.3030303 will be replaced by 0.6969697 which can be achieved be
## interchanging the sample labels!

## Warning in AUC(pred, group = as.integer(truth == namePos)): The computed
## AUC value 0.3030303 will be replaced by 0.6969697 which can be achieved be
## interchanging the sample labels!
## 
##      Performance Score(s)
## 
##                        Score     Value
## 1     area under curve (AUC) 0.6969697
## 2          Gini index (GINI) 0.3939394
## 3           Brier score (BS) 0.3658052
## 4 positive Brier score (PBS) 0.2336342
## 5 negative Brier score (NBS) 0.4499141
## 6 weighted Brier score (WBS) 0.3417741
## 7 balanced Brier score (BBS) 0.3417741
optCutoff(preds, truth =  as.numeric(datiscs$resdair), namePos = 1)
## Optimal Cut-off             YJS 
##       0.2389163       0.0000000

4.3 bondad de ajuste Hosmer Lemeshow, grafico calibración

 HLgof.test(fit = preds, obs = as.numeric(datiscs$resdair))
## $C
## 
##  Hosmer-Lemeshow C statistic
## 
## data:  preds and as.numeric(datiscs$resdair)
## X-squared = 272.7, df = 8, p-value < 2.2e-16
## 
## 
## $H
## 
##  Hosmer-Lemeshow H statistic
## 
## data:  preds and as.numeric(datiscs$resdair)
## X-squared = 272.2, df = 8, p-value < 2.2e-16
plot(cfs, las=1)

## 
## n=54   Mean absolute error=0.035   Mean squared error=0.00156
## 0.9 Quantile of absolute error=0.067

Conclusion no hay buen ajuste-calibración (p<0.05)

4.4 klic_score

4.4.1 Performance Measure

perfMeasures(predk, truth = as.numeric(datisck$resdair), namePos = 1)
## 
##      Performance Measure(s)
## 
##                                        Measure        Value
## 1                               accuracy (ACC)  0.414634146
## 2  probability of correct classification (PCC)  0.414634146
## 3                        fraction correct (FC)  0.414634146
## 4            simple matching coefficient (SMC)  0.414634146
## 5                Rand (similarity) index (RSI)  0.414634146
## 6       probability of misclassification (PMC)  0.585365854
## 7                              error rate (ER)  0.585365854
## 8                     fraction incorrect (FIC)  0.585365854
## 9                           sensitivity (SENS)  0.687500000
## 10                                recall (REC)  0.687500000
## 11                    true positive rate (TPR)  0.687500000
## 12               probability of detection (PD)  0.687500000
## 13                               hit rate (HR)  0.687500000
## 14                          specificity (SPEC)  0.240000000
## 15                    true negative rate (TNR)  0.240000000
## 16                           selectivity (SEL)  0.240000000
## 17                         detection rate (DR)  0.268292683
## 18                   false positive rate (FPR)  0.760000000
## 19                               fall-out (FO)  0.760000000
## 20                    false alarm (rate) (FAR)  0.760000000
## 21            probability of false alarm (PFA)  0.760000000
## 22                   false negative rate (FNR)  0.312500000
## 23                              miss rate (MR)  0.312500000
## 24                  false discovery rate (FDR)  0.633333333
## 25                   false omission rate (FOR)  0.454545455
## 26                           prevalence (PREV)  0.390243902
## 27      (positive) pre-test probability (PREP)  0.390243902
## 28             (positive) pre-test odds (PREO)  0.640000000
## 29                detection prevalence (DPREV)  0.731707317
## 30       negative pre-test probability (NPREP)  0.609756098
## 31              negative pre-test odds (NPREO)  1.562500000
## 32                   no information rate (NIR)  0.609756098
## 33                    weighted accuracy (WACC)  0.463750000
## 34                    balanced accuracy (BACC)  0.463750000
## 35              (bookmaker) informedness (INF) -0.072500000
## 36                  Youden's J statistic (YJS) -0.072500000
## 37                               deltap' (DPp) -0.072500000
## 38             positive likelihood ratio (PLR)  0.904605263
## 39             negative likelihood ratio (NLR)  1.302083333
## 40             weighted likelihood ratio (WLR)  1.103344298
## 41             balanced likelihood ratio (BLR)  1.103344298
## 42                 diagnostic odds ratio (DOR)  0.694736842
## 43             positive predictive value (PPV)  0.366666667
## 44                            precision (PREC)  0.366666667
## 45    (positive) post-test probability (POSTP)  0.366666667
## 46           (positive) post-test odds (POSTO)  0.578947368
## 47                      Bayes factor G1 (BFG1)  0.904605263
## 48             negative predictive value (NPV)  0.545454545
## 49     negative post-test probability (NPOSTP)  0.545454545
## 50            negative post-test odds (NPOSTO)  1.200000000
## 51                      Bayes factor G0 (BFG0)  0.768000000
## 52                           markedness (MARK) -0.087878788
## 53                                 deltap (DP) -0.087878788
## 54             weighted predictive value (WPV)  0.456060606
## 55             balanced predictive value (BPV)  0.456060606
## 56                              F1 score (F1S)  0.478260870
## 57           Dice similarity coefficient (DSC)  0.478260870
## 58                          F beta score (FBS)  0.478260870
## 59        Jaccard similarity coefficient (JSC)  0.314285714
## 60                           threat score (TS)  0.314285714
## 61                critical success index (CSI)  0.314285714
## 62     Matthews' correlation coefficient (MCC) -0.079819873
## 63        Pearson's correlation (r phi) (RPHI) -0.079819873
## 64                      Phi coefficient (PHIC) -0.079819873
## 65                            Cramer's V (CRV) -0.079819873
## 66    proportion of positive predictions (PPP)  0.731707317
## 67                    expected accuracy (EACC)  0.449137418
## 68             Cohen's kappa coefficient (CKC) -0.062634989
## 69            mutual information in bits (MI2)  0.004549947
## 70                 joint entropy in bits (JE2)  1.247256579
## 71      variation of information in bits (VI2)  1.794860935
## 72                       Jaccard distance (JD)  0.997471424
## 73           information quality ratio (INFQR)  0.002528576
## 74                uncertainty coefficient (UC)  0.004715182
## 75                    entropy coefficient (EC)  0.004715182
## 76                 proficiency (metric) (PROF)  0.004715182
## 77                   deficiency (metric) (DFM)  0.995284818
## 78                            redundancy (RED)  0.002522198
## 79                  symmetric uncertainty (SU)  0.005044396
## 80                 normalized uncertainty (NU)  0.005056737

4.4.2 Performance Score and optimal cutoff

perfScores(predk, truth = as.numeric(datisck$resdair), namePos = 1)
## Warning in AUC(pred, group = as.integer(truth == namePos)): The computed AUC
## value 0.32125 will be replaced by 0.67875 which can be achieved be interchanging
## the sample labels!

## Warning in AUC(pred, group = as.integer(truth == namePos)): The computed AUC
## value 0.32125 will be replaced by 0.67875 which can be achieved be interchanging
## the sample labels!
## 
##      Performance Score(s)
## 
##                        Score     Value
## 1     area under curve (AUC) 0.6787500
## 2          Gini index (GINI) 0.3575000
## 3           Brier score (BS) 0.3626403
## 4 positive Brier score (PBS) 0.2332582
## 5 negative Brier score (NBS) 0.4454449
## 6 weighted Brier score (WBS) 0.3393515
## 7 balanced Brier score (BBS) 0.3393515
optCutoff(predk, truth =  as.numeric(datisck$resdair), namePos = 1)
## Optimal Cut-off             YJS 
##       0.2142107       0.0000000

4.5 bondad de ajuste Hosmer Lemeshow, grafico calibración

 HLgof.test(fit = predk, obs = as.numeric(datisck$resdair))
## Warning in HLgof.test(fit = predk, obs = as.numeric(datisck$resdair)): Found
## only 6 different groups for Hosmer-Lemesho C statistic.
## $C
## 
##  Hosmer-Lemeshow C statistic
## 
## data:  predk and as.numeric(datisck$resdair)
## X-squared = 211.41, df = 4, p-value < 2.2e-16
## 
## 
## $H
## 
##  Hosmer-Lemeshow H statistic
## 
## data:  predk and as.numeric(datisck$resdair)
## X-squared = 211.51, df = 8, p-value < 2.2e-16
plot(cfk, las=1)

## 
## n=41   Mean absolute error=0.162   Mean squared error=0.04817
## 0.9 Quantile of absolute error=0.219

Conclusion no hay buen ajuste-calibración (p<0.05)

5 comparacion capacidad de discriminación klic-shohat

 AUC.test(pred1 = xs, lab1 = g, pred2 = xk, lab2 = g)
## Warning in AUC(pred2, group = lab2): The computed AUC value 0.32125 will be
## replaced by 0.67875 which can be achieved be interchanging the sample labels!
## $Variable1
##        AUC         SE     low CI      up CI 
## 0.69696970 0.06577784 0.56804750 0.82589190 
## 
## $Variable2
##        AUC         SE     low CI      up CI 
## 0.67875000 0.06724288 0.54695638 0.81054362 
## 
## $Test
## 
##  Hanley and McNeil test for two AUCs
## 
## data:  xs and xk
## z = 0.19369, p-value = 0.8464
## alternative hypothesis: true difference in AUC is not equal to 0
## 95 percent confidence interval:
##  -0.1661454  0.2025848
## sample estimates:
## Difference in AUC 
##         0.0182197

Conclusion no hay diferencia entre klic-shohat en discriminacion