Analisis3
2025-11-21
Análisis Exploratorio
library(readxl)
PBI_LPI <- read_excel("PBI_LPI.xls")
PBI_LPI## # A tibble: 160 × 14
## N PBIPEA Agr Ind Comercio Exp_Mer Exp_ByS Exp_Comb Exp_MM
## <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 1 6276. 0.210 0.113 0.366 0.00901 0.0690 0 0.000101
## 2 2 28603. 0.199 0.0566 0.558 0.0515 0.289 0.00576 0.00454
## 3 3 52937. 0.123 0.0561 0.485 0.0475 0.210 0.0447 0.000120
## 4 4 15749. 0 0 0.418 0.146 0.300 0 0
## 5 5 45039. 0.0635 0.136 0.205 0.0652 0.126 0.00165 0.00218
## 6 6 20792. 0.164 0.103 0.480 0.0614 0.331 0.00235 0.0233
## 7 7 89515. 0.0243 0.0607 0.322 0.169 0.193 0.0467 0.0535
## 8 8 90843. 0.0110 0.162 0.793 0.369 0.523 0.00630 0.0118
## 9 9 33202. 0.056 0.049 0.572 0.0796 0.464 0.0696 0.00129
## 10 10 40889. 0.00880 0.0316 0.265 0.0500 0.358 0 0
## # ℹ 150 more rows
## # ℹ 5 more variables: Prod_oil <dbl>, RRNN <dbl>, Tierras_Agr <dbl>,
## # Exp_H_Tech <dbl>, LPI <dbl>plot (PBI_LPI)
library(car)## Loading required package: carDataPBI_LPI_1 <- PBI_LPI[ ,c(2,3,4,5,6,7,8,9,10,11,12,13,14)]
PBI_LPI_1## # A tibble: 160 × 13
## PBIPEA Agr Ind Comercio Exp_Mer Exp_ByS Exp_Comb Exp_MM Prod_oil
## <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 6276. 0.210 0.113 0.366 0.00901 0.0690 0 0.000101 0
## 2 28603. 0.199 0.0566 0.558 0.0515 0.289 0.00576 0.00454 171.
## 3 52937. 0.123 0.0561 0.485 0.0475 0.210 0.0447 0.000120 1129.
## 4 15749. 0 0 0.418 0.146 0.300 0 0 1491.
## 5 45039. 0.0635 0.136 0.205 0.0652 0.126 0.00165 0.00218 259.
## 6 20792. 0.164 0.103 0.480 0.0614 0.331 0.00235 0.0233 0
## 7 89515. 0.0243 0.0607 0.322 0.169 0.193 0.0467 0.0535 227.
## 8 90843. 0.0110 0.162 0.793 0.369 0.523 0.00630 0.0118 33.4
## 9 33202. 0.056 0.049 0.572 0.0796 0.464 0.0696 0.00129 1678.
## 10 40889. 0.00880 0.0316 0.265 0.0500 0.358 0 0 0
## # ℹ 150 more rows
## # ℹ 4 more variables: RRNN <dbl>, Tierras_Agr <dbl>, Exp_H_Tech <dbl>,
## # LPI <dbl>scatterplotMatrix(PBI_LPI_1)
library(corrplot)## corrplot 0.95 loadedM <- cor(PBI_LPI_1)
corrplot(M)
library(psych)##
## Attaching package: 'psych'## The following object is masked from 'package:car':
##
## logitcorPlot(PBI_LPI_1)
## Componentes
PC1 <- princomp(PBI_LPI_1)
PC1## Call:
## princomp(x = PBI_LPI_1)
##
## Standard deviations:
## Comp.1 Comp.2 Comp.3 Comp.4 Comp.5 Comp.6
## 3.874850e+04 1.430167e+03 4.528068e-01 3.510142e-01 2.052192e-01 1.389599e-01
## Comp.7 Comp.8 Comp.9 Comp.10 Comp.11 Comp.12
## 9.436348e-02 7.346306e-02 5.800053e-02 5.002169e-02 2.268443e-02 1.699327e-02
## Comp.13
## 1.401994e-02
##
## 13 variables and 160 observations.plot(PC1)
summary(PC1)## Importance of components:
## Comp.1 Comp.2 Comp.3 Comp.4
## Standard deviation 3.874850e+04 1.430167e+03 4.528068e-01 3.510142e-01
## Proportion of Variance 9.986396e-01 1.360419e-03 1.363719e-10 8.194992e-11
## Cumulative Proportion 9.986396e-01 1.000000e+00 1.000000e+00 1.000000e+00
## Comp.5 Comp.6 Comp.7 Comp.8
## Standard deviation 2.052192e-01 1.389599e-01 9.436348e-02 7.346306e-02
## Proportion of Variance 2.801141e-11 1.284334e-11 5.922526e-12 3.589526e-12
## Cumulative Proportion 1.000000e+00 1.000000e+00 1.000000e+00 1.000000e+00
## Comp.9 Comp.10 Comp.11 Comp.12
## Standard deviation 5.800053e-02 5.002169e-02 2.268443e-02 1.699327e-02
## Proportion of Variance 2.237500e-12 1.664240e-12 3.422590e-13 1.920671e-13
## Cumulative Proportion 1.000000e+00 1.000000e+00 1.000000e+00 1.000000e+00
## Comp.13
## Standard deviation 1.401994e-02
## Proportion of Variance 1.307348e-13
## Cumulative Proportion 1.000000e+00biplot(PC1)## Warning in arrows(0, 0, y[, 1L] * 0.8, y[, 2L] * 0.8, col = col[2L], length =
## arrow.len): zero-length arrow is of indeterminate angle and so skipped
## Warning in arrows(0, 0, y[, 1L] * 0.8, y[, 2L] * 0.8, col = col[2L], length =
## arrow.len): zero-length arrow is of indeterminate angle and so skipped
## Warning in arrows(0, 0, y[, 1L] * 0.8, y[, 2L] * 0.8, col = col[2L], length =
## arrow.len): zero-length arrow is of indeterminate angle and so skipped
## Warning in arrows(0, 0, y[, 1L] * 0.8, y[, 2L] * 0.8, col = col[2L], length =
## arrow.len): zero-length arrow is of indeterminate angle and so skipped
## Warning in arrows(0, 0, y[, 1L] * 0.8, y[, 2L] * 0.8, col = col[2L], length =
## arrow.len): zero-length arrow is of indeterminate angle and so skipped
## Warning in arrows(0, 0, y[, 1L] * 0.8, y[, 2L] * 0.8, col = col[2L], length =
## arrow.len): zero-length arrow is of indeterminate angle and so skipped
## Warning in arrows(0, 0, y[, 1L] * 0.8, y[, 2L] * 0.8, col = col[2L], length =
## arrow.len): zero-length arrow is of indeterminate angle and so skipped
## Warning in arrows(0, 0, y[, 1L] * 0.8, y[, 2L] * 0.8, col = col[2L], length =
## arrow.len): zero-length arrow is of indeterminate angle and so skipped
## Warning in arrows(0, 0, y[, 1L] * 0.8, y[, 2L] * 0.8, col = col[2L], length =
## arrow.len): zero-length arrow is of indeterminate angle and so skipped
## Warning in arrows(0, 0, y[, 1L] * 0.8, y[, 2L] * 0.8, col = col[2L], length =
## arrow.len): zero-length arrow is of indeterminate angle and so skipped
## Warning in arrows(0, 0, y[, 1L] * 0.8, y[, 2L] * 0.8, col = col[2L], length =
## arrow.len): zero-length arrow is of indeterminate angle and so skipped
acp2 <-PC1$scores[ ,1:2]
plot(acp2)
Clusters
library(cluster)
agp3 = agnes(acp2,method="average")
plot(agp3)
agpcut <- cutree(agp3,3)
par(mfrow=c(1,1))
plot(acp2,col=agpcut)
Clusters Clara
plot(clara(PBI_LPI_1,6))
modelo <- clara(PBI_LPI_1, k = 6)
modelo$clustering## [1] 1 2 3 2 4 2 5 5 4 4 5 1 4 5 1 2 2 4 4 4 6 4 1 1 1 1 5 1 3 2 2 1 1 2 4 1 3
## [38] 3 3 5 1 4 2 4 2 1 3 1 2 5 5 3 1 2 5 1 3 2 1 1 2 1 1 5 3 3 2 2 3 3 6 5 5 2
## [75] 3 4 4 1 3 4 6 1 1 3 4 1 1 3 6 1 1 3 2 1 5 2 4 4 2 4 4 2 1 4 1 5 3 1 1 2 6
## [112] 5 2 4 1 2 2 2 3 3 6 3 3 1 1 6 1 4 1 6 3 3 1 1 4 5 2 5 5 1 1 1 2 1 3 4 3 4
## [149] 1 2 5 5 5 4 2 4 1 1 1 1clusters <- split(1:nrow(PBI_LPI_1), modelo$clustering)
clusters## $`1`
## [1] 1 12 15 23 24 25 26 28 32 33 36 41 46 48 53 56 59 60 62
## [20] 63 78 82 83 86 87 90 91 94 103 105 108 109 115 124 125 127 129 133
## [39] 134 140 141 142 144 149 157 158 159 160
##
## $`2`
## [1] 2 4 6 16 17 30 31 34 43 45 49 54 58 61 67 68 74 93 96
## [20] 99 102 110 113 116 117 118 137 143 150 155
##
## $`3`
## [1] 3 29 37 38 39 47 52 57 65 66 69 70 75 79 84 88 92 107 119
## [20] 120 122 123 131 132 145 147
##
## $`4`
## [1] 5 9 10 13 18 19 20 22 35 42 44 76 77 80 85 97 98 100 101
## [20] 104 114 128 135 146 148 154 156
##
## $`5`
## [1] 7 8 11 14 27 40 50 51 55 64 72 73 95 106 112 136 138 139 151
## [20] 152 153
##
## $`6`
## [1] 21 71 81 89 111 121 126 130PBI_LPI_cluster<- PBI_LPI
PBI_LPI_cluster$cluster <- modelo$clustering
PBI_LPI_cluster## # A tibble: 160 × 15
## N PBIPEA Agr Ind Comercio Exp_Mer Exp_ByS Exp_Comb Exp_MM
## <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 1 6276. 0.210 0.113 0.366 0.00901 0.0690 0 0.000101
## 2 2 28603. 0.199 0.0566 0.558 0.0515 0.289 0.00576 0.00454
## 3 3 52937. 0.123 0.0561 0.485 0.0475 0.210 0.0447 0.000120
## 4 4 15749. 0 0 0.418 0.146 0.300 0 0
## 5 5 45039. 0.0635 0.136 0.205 0.0652 0.126 0.00165 0.00218
## 6 6 20792. 0.164 0.103 0.480 0.0614 0.331 0.00235 0.0233
## 7 7 89515. 0.0243 0.0607 0.322 0.169 0.193 0.0467 0.0535
## 8 8 90843. 0.0110 0.162 0.793 0.369 0.523 0.00630 0.0118
## 9 9 33202. 0.056 0.049 0.572 0.0796 0.464 0.0696 0.00129
## 10 10 40889. 0.00880 0.0316 0.265 0.0500 0.358 0 0
## # ℹ 150 more rows
## # ℹ 6 more variables: Prod_oil <dbl>, RRNN <dbl>, Tierras_Agr <dbl>,
## # Exp_H_Tech <dbl>, LPI <dbl>, cluster <int>library(openxlsx)
write.xlsx(PBI_LPI_cluster, file = "PBI_LPI_cluster")RED NEURONAL
library(neuralnet) # regression
library(nnet) # classification
library(NeuralNetTools)
library(plyr)normalize<-function(x){
return ( (x-min(x))/(max(x)-min(x)))
}
PBI_LPI_norm<-as.data.frame(lapply(PBI_LPI_1,FUN=normalize))
summary(PBI_LPI_norm)## PBIPEA Agr Ind Comercio
## Min. :0.00000 Min. :0.00000 Min. :0.0000 Min. :0.0000
## 1st Qu.:0.03572 1st Qu.:0.03331 1st Qu.:0.1884 1st Qu.:0.1191
## Median :0.13322 Median :0.10551 Median :0.3406 Median :0.1621
## Mean :0.18355 Mean :0.17648 Mean :0.3332 Mean :0.1881
## 3rd Qu.:0.28131 3rd Qu.:0.27452 3rd Qu.:0.4560 3rd Qu.:0.2189
## Max. :1.00000 Max. :1.00000 Max. :1.0000 Max. :1.0000
## Exp_Mer Exp_ByS Exp_Comb Exp_MM
## Min. :0.00000 Min. :0.0000 Min. :0.000e+00 Min. :0.00000
## 1st Qu.:0.04445 1st Qu.:0.0946 1st Qu.:1.360e-06 1st Qu.:0.00123
## Median :0.07775 Median :0.1454 Median :1.306e-02 Median :0.01673
## Mean :0.11938 Mean :0.1752 Mean :9.673e-02 Mean :0.05765
## 3rd Qu.:0.14587 3rd Qu.:0.2088 3rd Qu.:8.600e-02 3rd Qu.:0.04817
## Max. :1.00000 Max. :1.0000 Max. :1.000e+00 Max. :1.00000
## Prod_oil RRNN Tierras_Agr Exp_H_Tech
## Min. :0.000e+00 Min. :0.00000 Min. :0.0000 Min. :0.000000
## 1st Qu.:0.000e+00 1st Qu.:0.01106 1st Qu.:0.3061 1st Qu.:0.000000
## Median :8.173e-05 Median :0.04794 Median :0.5063 Median :0.008377
## Mean :3.789e-02 Mean :0.13455 Mean :0.4991 Mean :0.048017
## 3rd Qu.:7.798e-03 3rd Qu.:0.21329 3rd Qu.:0.7103 3rd Qu.:0.046164
## Max. :1.000e+00 Max. :1.00000 Max. :1.0000 Max. :1.000000
## LPI
## Min. :0.0000
## 1st Qu.:0.2962
## Median :0.4038
## Mean :0.4644
## 3rd Qu.:0.5977
## Max. :1.0000indice <- sample(2, nrow(PBI_LPI_norm), replace = TRUE, prob = c(0.7,0.3))
PBI_LPI_train <- PBI_LPI_1[indice==1,]
PBI_LPI_test <- PBI_LPI_1[indice==2,]library(neuralnet)
PBI_LPI_model <- neuralnet(LPI ~ Agr + Ind + Comercio + Exp_Mer + Exp_ByS + Exp_Comb + Exp_MM + Prod_oil + RRNN + Tierras_Agr + Exp_H_Tech , data = PBI_LPI_train)
str(PBI_LPI_model)## List of 14
## $ call : language neuralnet(formula = LPI ~ Agr + Ind + Comercio + Exp_Mer + Exp_ByS + Exp_Comb + Exp_MM + Prod_oil + RRNN + T| __truncated__
## $ response : num [1:116, 1] 2.13 2.51 2.16 2.93 2.49 ...
## ..- attr(*, "dimnames")=List of 2
## .. ..$ : chr [1:116] "1" "2" "3" "4" ...
## .. ..$ : chr "LPI"
## $ covariate : num [1:116, 1:11] 0.2097 0.1991 0 0.0635 0.1639 ...
## ..- attr(*, "dimnames")=List of 2
## .. ..$ : NULL
## .. ..$ : chr [1:11] "Agr" "Ind" "Comercio" "Exp_Mer" ...
## $ model.list :List of 2
## ..$ response : chr "LPI"
## ..$ variables: chr [1:11] "Agr" "Ind" "Comercio" "Exp_Mer" ...
## $ err.fct :function (x, y)
## ..- attr(*, "type")= chr "sse"
## $ act.fct :function (x)
## ..- attr(*, "type")= chr "logistic"
## $ linear.output : logi TRUE
## $ data :'data.frame': 116 obs. of 13 variables:
## ..$ PBIPEA : num [1:116] 6276 28603 15749 45039 20792 ...
## ..$ Agr : num [1:116] 0.2097 0.1991 0 0.0635 0.1639 ...
## ..$ Ind : num [1:116] 0.1131 0.0566 0 0.136 0.1028 ...
## ..$ Comercio : num [1:116] 0.366 0.558 0.418 0.205 0.48 ...
## ..$ Exp_Mer : num [1:116] 0.00901 0.05151 0.14606 0.06522 0.06142 ...
## ..$ Exp_ByS : num [1:116] 0.069 0.289 0.3 0.126 0.331 ...
## ..$ Exp_Comb : num [1:116] 0 0.00576 0 0.00165 0.00235 ...
## ..$ Exp_MM : num [1:116] 0.000101 0.004538 0 0.002181 0.023324 ...
## ..$ Prod_oil : num [1:116] 0 171 1491 259 0 ...
## ..$ RRNN : num [1:116] 0.00598 0.01973 0.13217 0.01302 0.04827 ...
## ..$ Tierras_Agr: num [1:116] 0.581 0.429 0.475 0.543 0.589 ...
## ..$ Exp_H_Tech : num [1:116] 0 0.000334 0 0.005735 0.003648 ...
## ..$ LPI : num [1:116] 2.13 2.51 2.16 2.93 2.49 ...
## $ exclude : NULL
## $ net.result :List of 1
## ..$ : num [1:116, 1] 2.37 2.37 2.45 2.87 2.42 ...
## $ weights :List of 1
## ..$ :List of 2
## .. ..$ : num [1:12, 1] -2.27 -22.52 18.79 -4.12 8.5 ...
## .. ..$ : num [1:2, 1] 2.37 1.51
## $ generalized.weights:List of 1
## ..$ : num [1:116, 1:11] 0.0264 0.0139 0.5067 1.4066 0.3229 ...
## $ startweights :List of 1
## ..$ :List of 2
## .. ..$ : num [1:12, 1] -0.4337 0.4647 0.0253 0.6412 -2.2035 ...
## .. ..$ : num [1:2, 1] -0.8486 0.0625
## $ result.matrix : num [1:17, 1] 4.5789 0.0041 1808 -2.2728 -22.5195 ...
## ..- attr(*, "dimnames")=List of 2
## .. ..$ : chr [1:17] "error" "reached.threshold" "steps" "Intercept.to.1layhid1" ...
## .. ..$ : NULL
## - attr(*, "class")= chr "nn"plot(PBI_LPI_model, rep = "best")
model_results <- compute(PBI_LPI_model,PBI_LPI_test[1:13])
predicted_LPI <- model_results$net.resultcor(predicted_LPI,PBI_LPI_test$LPI)## [,1]
## [1,] 0.7829385predicted_LPI## [,1]
## [1,] 2.370268
## [2,] 2.374159
## [3,] 2.588066
## [4,] 3.843029
## [5,] 2.469517
## [6,] 2.370408
## [7,] 2.545683
## [8,] 3.808853
## [9,] 2.508531
## [10,] 2.405733
## [11,] 2.443978
## [12,] 2.429153
## [13,] 2.456763
## [14,] 2.618534
## [15,] 3.441217
## [16,] 2.369351
## [17,] 2.516894
## [18,] 2.775076
## [19,] 2.567271
## [20,] 3.493072
## [21,] 2.456106
## [22,] 2.747431
## [23,] 3.861237
## [24,] 2.369089
## [25,] 2.368942
## [26,] 2.743455
## [27,] 2.458317
## [28,] 2.872181
## [29,] 2.378721
## [30,] 2.373238
## [31,] 2.588420
## [32,] 2.405902
## [33,] 3.509258
## [34,] 2.895118
## [35,] 3.352965
## [36,] 2.368888
## [37,] 2.379451
## [38,] 2.368591
## [39,] 2.377482
## [40,] 2.390365
## [41,] 2.368663
## [42,] 2.368776
## [43,] 3.436184
## [44,] 3.370929PBI_LPI_test$LPI## [1] 2.455828 2.254286 2.759056 3.180760 1.937576 2.359471 2.585589 3.026809
## [9] 2.753272 2.331307 2.045015 2.630513 3.003740 2.097716 3.814949 2.380741
## [17] 2.466700 2.896920 2.882424 3.695022 2.389892 2.622612 3.820868 2.364158
## [25] 2.027063 2.527753 2.738210 3.390797 2.478651 2.663394 2.929829 2.300339
## [33] 2.791576 2.931236 3.353241 2.354357 2.327711 1.785748 2.389656 1.642313
## [41] 2.255333 2.528847 3.980922 3.966106Mejorar el modelo
PBI_LPI_model2 <- neuralnet(LPI ~ Agr + Ind + Comercio + Exp_Mer + Exp_ByS + Exp_Comb + Exp_MM + Prod_oil + RRNN + Tierras_Agr + Exp_H_Tech , data = PBI_LPI_train, hidden = c(2,4))
plot(PBI_LPI_model2 ,rep = "best")
model_results2 <- compute(PBI_LPI_model2,PBI_LPI_test[1:13])
predicted_LPI2 <- model_results2$net.result
cor(predicted_LPI2,PBI_LPI_test$LPI)## [,1]
## [1,] 0.3678225predicted_LPI2## [,1]
## [1,] 2.251338
## [2,] 2.262804
## [3,] 3.795465
## [4,] 3.763839
## [5,] 2.482329
## [6,] 2.298121
## [7,] 2.925812
## [8,] 3.358604
## [9,] 2.540178
## [10,] 3.229286
## [11,] 3.687699
## [12,] 2.591372
## [13,] 2.586871
## [14,] 3.241967
## [15,] 3.218938
## [16,] 2.272910
## [17,] 3.353442
## [18,] 2.737070
## [19,] 2.952647
## [20,] 3.290148
## [21,] 2.528378
## [22,] 2.917347
## [23,] 3.727752
## [24,] 2.306503
## [25,] 2.263656
## [26,] 2.978076
## [27,] 2.441635
## [28,] 2.829389
## [29,] 2.357855
## [30,] 2.648245
## [31,] 2.942413
## [32,] 2.317951
## [33,] 3.113339
## [34,] 3.012232
## [35,] 3.091225
## [36,] 2.356732
## [37,] 2.361321
## [38,] 2.266459
## [39,] 3.335811
## [40,] 3.635318
## [41,] 2.423627
## [42,] 2.354919
## [43,] 3.144276
## [44,] 2.957650PBI_LPI_test$LPI## [1] 2.455828 2.254286 2.759056 3.180760 1.937576 2.359471 2.585589 3.026809
## [9] 2.753272 2.331307 2.045015 2.630513 3.003740 2.097716 3.814949 2.380741
## [17] 2.466700 2.896920 2.882424 3.695022 2.389892 2.622612 3.820868 2.364158
## [25] 2.027063 2.527753 2.738210 3.390797 2.478651 2.663394 2.929829 2.300339
## [33] 2.791576 2.931236 3.353241 2.354357 2.327711 1.785748 2.389656 1.642313
## [41] 2.255333 2.528847 3.980922 3.966106