Transformation non-normal distribution to the normal distribution by using the method of log transformation, BOX COX transformation and randomly ordered.
Data Analysis :
data <- read.csv("C:/Users/gaming/OneDrive/Desktop/Dena Qais/karpur (2).csv",header = TRUE) #calling data
dim(data)
## [1] 819 15
summary(data)
## depth caliper ind.deep ind.med
## Min. :5667 Min. :8.487 Min. : 6.532 Min. : 9.386
## 1st Qu.:5769 1st Qu.:8.556 1st Qu.: 28.799 1st Qu.: 27.892
## Median :5872 Median :8.588 Median :217.849 Median :254.383
## Mean :5873 Mean :8.622 Mean :275.357 Mean :273.357
## 3rd Qu.:5977 3rd Qu.:8.686 3rd Qu.:566.793 3rd Qu.:544.232
## Max. :6083 Max. :8.886 Max. :769.484 Max. :746.028
## gamma phi.N R.deep R.med
## Min. : 16.74 Min. :0.0150 Min. : 1.300 Min. : 1.340
## 1st Qu.: 40.89 1st Qu.:0.2030 1st Qu.: 1.764 1st Qu.: 1.837
## Median : 51.37 Median :0.2450 Median : 4.590 Median : 3.931
## Mean : 53.42 Mean :0.2213 Mean : 24.501 Mean : 21.196
## 3rd Qu.: 62.37 3rd Qu.:0.2640 3rd Qu.: 34.724 3rd Qu.: 35.853
## Max. :112.40 Max. :0.4100 Max. :153.085 Max. :106.542
## SP density.corr density phi.core
## Min. :-73.95 Min. :-0.067000 Min. :1.758 Min. :15.70
## 1st Qu.:-42.01 1st Qu.:-0.016000 1st Qu.:2.023 1st Qu.:23.90
## Median :-32.25 Median :-0.007000 Median :2.099 Median :27.60
## Mean :-30.98 Mean :-0.008883 Mean :2.102 Mean :26.93
## 3rd Qu.:-19.48 3rd Qu.: 0.002000 3rd Qu.:2.181 3rd Qu.:30.70
## Max. : 25.13 Max. : 0.089000 Max. :2.387 Max. :36.30
## k.core Facies phi.core.frac
## Min. : 0.42 Length:819 Min. :0.1570
## 1st Qu.: 657.33 Class :character 1st Qu.:0.2390
## Median : 1591.22 Mode :character Median :0.2760
## Mean : 2251.91 Mean :0.2693
## 3rd Qu.: 3046.82 3rd Qu.:0.3070
## Max. :15600.00 Max. :0.3630
head(data)
## depth caliper ind.deep ind.med gamma phi.N R.deep R.med SP
## 1 5667.0 8.685 618.005 569.781 98.823 0.410 1.618 1.755 -56.587
## 2 5667.5 8.686 497.547 419.494 90.640 0.307 2.010 2.384 -61.916
## 3 5668.0 8.686 384.935 300.155 78.087 0.203 2.598 3.332 -55.861
## 4 5668.5 8.686 278.324 205.224 66.232 0.119 3.593 4.873 -41.860
## 5 5669.0 8.686 183.743 131.155 59.807 0.069 5.442 7.625 -34.934
## 6 5669.5 8.686 109.512 75.633 57.109 0.048 9.131 13.222 -39.769
## density.corr density phi.core k.core Facies phi.core.frac
## 1 -0.033 2.205 33.9000 2442.590 F1 0.339000
## 2 -0.067 2.040 33.4131 3006.989 F1 0.334131
## 3 -0.064 1.888 33.1000 3370.000 F1 0.331000
## 4 -0.053 1.794 34.9000 2270.000 F1 0.349000
## 5 -0.054 1.758 35.0644 2530.758 F1 0.350644
## 6 -0.058 1.759 35.3152 2928.314 F1 0.353152
tail(data)
## depth caliper ind.deep ind.med gamma phi.N R.deep R.med SP
## 814 6080.5 8.578 683.847 672.125 24.003 0.208 1.462 1.488 -33.714
## 815 6081.0 8.590 678.002 669.171 27.855 0.214 1.475 1.494 -4.808
## 816 6081.5 8.588 668.633 661.949 32.591 0.228 1.496 1.511 -4.727
## 817 6082.0 8.588 656.328 647.918 38.547 0.243 1.524 1.543 -21.390
## 818 6082.5 8.588 643.216 628.536 45.555 0.256 1.555 1.591 -31.597
## 819 6083.0 8.588 631.098 608.163 52.244 0.265 1.584 1.644 -36.109
## density.corr density phi.core k.core Facies phi.core.frac
## 814 0.003 2.147 26.9000 1300.790 F5 0.269000
## 815 -0.002 2.162 25.7547 1249.706 F5 0.257547
## 816 -0.008 2.158 24.5569 1196.282 F5 0.245569
## 817 -0.006 2.136 23.3592 1142.859 F5 0.233592
## 818 -0.002 2.115 22.1614 1089.435 F5 0.221614
## 819 -0.003 2.109 20.9637 1036.012 F5 0.209637
range(data$depth)
## [1] 5667 6083
histogram of core permeability and core porosity :
par(mfrow=c(1,2))
hist(data$k.core, main='Histogram of Permeability', xlab='Permeability, md', col='darkred')
hist(data$phi.core, main='Histogram of Porosity', xlab='Porosity', col='darkgreen')
Multiple Linear Regression of Core Permeability as a function of other data :
model1 <- lm(k.core ~ .-1, data = data) #k.core as function of all data
summary(model1)
##
## Call:
## lm(formula = k.core ~ . - 1, data = data)
##
## Residuals:
## Min 1Q Median 3Q Max
## -5585.6 -568.9 49.2 476.5 8928.4
##
## Coefficients: (1 not defined because of singularities)
## Estimate Std. Error t value Pr(>|t|)
## depth 8.544e+00 1.785e+00 4.786 2.02e-06 ***
## caliper 1.413e+03 1.019e+03 1.387 0.165789
## ind.deep -2.418e-01 2.354e+00 -0.103 0.918220
## ind.med 1.224e+00 2.585e+00 0.473 0.636062
## gamma -4.583e+01 6.010e+00 -7.626 6.88e-14 ***
## phi.N -2.010e+03 1.476e+03 -1.362 0.173540
## R.deep -2.344e+01 6.288e+00 -3.727 0.000207 ***
## R.med 5.643e+01 9.065e+00 6.225 7.76e-10 ***
## SP -7.125e+00 3.145e+00 -2.266 0.023736 *
## density.corr -2.567e+03 4.809e+03 -0.534 0.593602
## density 2.319e+03 1.173e+03 1.976 0.048458 *
## phi.core 1.921e+02 2.282e+01 8.418 < 2e-16 ***
## FaciesF1 -6.783e+04 1.760e+04 -3.853 0.000126 ***
## FaciesF10 -6.694e+04 1.765e+04 -3.793 0.000160 ***
## FaciesF2 -6.691e+04 1.761e+04 -3.800 0.000156 ***
## FaciesF3 -6.740e+04 1.767e+04 -3.815 0.000147 ***
## FaciesF5 -6.709e+04 1.767e+04 -3.798 0.000157 ***
## FaciesF7 -6.788e+04 1.764e+04 -3.848 0.000129 ***
## FaciesF8 -6.901e+04 1.770e+04 -3.900 0.000104 ***
## FaciesF9 -7.080e+04 1.779e+04 -3.980 7.52e-05 ***
## phi.core.frac NA NA NA NA
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1262 on 799 degrees of freedom
## Multiple R-squared: 0.8457, Adjusted R-squared: 0.8418
## F-statistic: 218.9 on 20 and 799 DF, p-value: < 2.2e-16
par(mfrow=c(2,2))
plot(model1)
Reduction the model1 by removing of none effective parameters that have value less than 0.05 ( less than the value of probability of neural hypothesis) pr(>|t|) :
model1.red <- lm(k.core ~ depth+gamma+R.deep+R.med+SP+density+phi.core+Facies-1, data = data)
summary(model1.red)
##
## Call:
## lm(formula = k.core ~ depth + gamma + R.deep + R.med + SP + density +
## phi.core + Facies - 1, data = data)
##
## Residuals:
## Min 1Q Median 3Q Max
## -5461.7 -545.5 37.0 505.0 9072.8
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## depth 8.088 1.059 7.639 6.23e-14 ***
## gamma -51.915 4.526 -11.471 < 2e-16 ***
## R.deep -22.467 6.261 -3.588 0.000353 ***
## R.med 48.859 8.730 5.597 2.99e-08 ***
## SP -6.490 3.118 -2.082 0.037671 *
## density 2013.607 1047.054 1.923 0.054818 .
## phi.core 188.002 21.791 8.628 < 2e-16 ***
## FaciesF1 -51703.918 5802.150 -8.911 < 2e-16 ***
## FaciesF10 -50893.923 5917.391 -8.601 < 2e-16 ***
## FaciesF2 -51009.565 5854.625 -8.713 < 2e-16 ***
## FaciesF3 -51322.213 5871.731 -8.741 < 2e-16 ***
## FaciesF5 -51174.044 6008.663 -8.517 < 2e-16 ***
## FaciesF7 -52302.690 5971.976 -8.758 < 2e-16 ***
## FaciesF8 -53362.629 6033.789 -8.844 < 2e-16 ***
## FaciesF9 -54796.231 6100.716 -8.982 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1266 on 804 degrees of freedom
## Multiple R-squared: 0.8436, Adjusted R-squared: 0.8407
## F-statistic: 289.1 on 15 and 804 DF, p-value: < 2.2e-16
par(mfrow=c(2,2))
plot(model1.red)
Prduction of K.core.red after the reduction data of model1 :
k.core.pred <- predict(model1.red,data)
Find the adjusted R- sequared that represent the indication of accuracy, and find Root main square error :
range(data$k.core)
## [1] 0.42 15600.00
AdjR.sq1 <- 1-sum((k.core.pred - data$k.core)^2)/sum((data$k.core-mean(data$k.core))^2)
AdjR.sq1
## [1] 0.6847518
mspe.model1 <- sqrt(sum((k.core.pred - data$k.core)^2)/nrow(data))
mspe.model1
## [1] 1254.46
plot the observed and Predicted values of K.core, plots between K.core, K.core.pred with depth.
par(mfrow=c(1,1))
plot(y=y<-(data$depth),ylim=rev(range(data$depth)),x=x<-(data$k.core),type="p", col="red", lwd = 5, pch=16, xlab='Permeability',
ylab='Depth, m', xlim=c(0,16000), cex=1.5, cex.lab=1.5, cex.axis=1.2)
grid()
par(new = TRUE)
plot(y=y<-(data$depth),ylim=rev(range(data$depth)),x=x<-(k.core.pred),type="l", col="blue", lwd = 5, pch=17, xlab='',
ylab='Depth, m', xlim=c(0,16000), cex=1.5, cex.lab=1.5, cex.axis=1.2, main='AdjR.sq1=0.6847 & RMSE=1254')
grid()
legend('topright', legend=c("Observed", "Predicted"), pch=c(16,17), col=c("red", "blue"))
Core Permeability Modeling by using Log Transforamtion method :
model2<-lm(log10(k.core) ~ .-1,data=data) #k.core as function of all data
summary(model2)
##
## Call:
## lm(formula = log10(k.core) ~ . - 1, data = data)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.5804 -0.1138 0.0322 0.1529 0.7384
##
## Coefficients: (1 not defined because of singularities)
## Estimate Std. Error t value Pr(>|t|)
## depth 0.0007425 0.0004718 1.574 0.1160
## caliper -0.4605945 0.2693103 -1.710 0.0876 .
## ind.deep -0.0007951 0.0006222 -1.278 0.2017
## ind.med 0.0007137 0.0006833 1.044 0.2966
## gamma -0.0091269 0.0015885 -5.746 1.30e-08 ***
## phi.N -1.7628155 0.3901024 -4.519 7.16e-06 ***
## R.deep -0.0025878 0.0016620 -1.557 0.1199
## R.med 0.0044073 0.0023960 1.839 0.0662 .
## SP -0.0016935 0.0008312 -2.037 0.0419 *
## density.corr 1.4462633 1.2712045 1.138 0.2556
## density 1.6148374 0.3100921 5.208 2.44e-07 ***
## phi.core 0.0948634 0.0060329 15.724 < 2e-16 ***
## FaciesF1 -2.3461877 4.6532000 -0.504 0.6143
## FaciesF10 -2.2675417 4.6652182 -0.486 0.6271
## FaciesF2 -2.3646211 4.6544047 -0.508 0.6116
## FaciesF3 -2.3769425 4.6697172 -0.509 0.6109
## FaciesF5 -2.2367684 4.6697432 -0.479 0.6321
## FaciesF7 -2.0650257 4.6623059 -0.443 0.6579
## FaciesF8 -2.4438111 4.6776859 -0.522 0.6015
## FaciesF9 -2.7023993 4.7023453 -0.575 0.5657
## phi.core.frac NA NA NA NA
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3335 on 799 degrees of freedom
## Multiple R-squared: 0.9891, Adjusted R-squared: 0.9888
## F-statistic: 3613 on 20 and 799 DF, p-value: < 2.2e-16
par(mfrow=c(2,2))
plot(model2)
model2.red <- lm(log10(k.core) ~ caliper+gamma+phi.N+SP+density+phi.core+Facies-1,data=data)
summary(model2.red)
##
## Call:
## lm(formula = log10(k.core) ~ caliper + gamma + phi.N + SP + density +
## phi.core + Facies - 1, data = data)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.60108 -0.11891 0.03253 0.15792 0.70925
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## caliper -0.840136 0.182354 -4.607 4.74e-06 ***
## gamma -0.010369 0.001337 -7.754 2.68e-14 ***
## phi.N -1.596764 0.359392 -4.443 1.01e-05 ***
## SP -0.001585 0.000793 -1.998 0.04600 *
## density 1.762145 0.287056 6.139 1.31e-09 ***
## phi.core 0.093920 0.005931 15.835 < 2e-16 ***
## FaciesF1 4.981486 1.783281 2.793 0.00534 **
## FaciesF10 5.101106 1.793029 2.845 0.00455 **
## FaciesF2 4.995149 1.796855 2.780 0.00556 **
## FaciesF3 4.967152 1.799799 2.760 0.00591 **
## FaciesF5 5.158638 1.760353 2.930 0.00348 **
## FaciesF7 5.272425 1.759300 2.997 0.00281 **
## FaciesF8 4.952389 1.753512 2.824 0.00486 **
## FaciesF9 4.687264 1.762588 2.659 0.00799 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3346 on 805 degrees of freedom
## Multiple R-squared: 0.9889, Adjusted R-squared: 0.9887
## F-statistic: 5128 on 14 and 805 DF, p-value: < 2.2e-16
par(mfrow=c(2,2))
plot(model2.red)
k.core.pred2 <- predict(model2.red,data) #prediction of k.core after reduction data of model2
k.core.pred2 <- 10^(k.core.pred2)
AdjR.sq2 <- 1-sum((k.core.pred2 - data$k.core)^2)/sum((data$k.core-mean(data$k.core))^2)
AdjR.sq2
## [1] 0.6366977
mspe.model2 <- sqrt(sum((k.core.pred2 - data$k.core)^2)/nrow(data))
mspe.model2
## [1] 1346.681
plot the observed and Predicted values K.core, plots between K.core, K.core.pred with depth after using Log Transforamtion method :
par(mfrow=c(1,1))
plot(y=y<-(data$depth),ylim=rev(range(data$depth)),x=x<-(data$k.core),type="p", col="red", lwd = 5, pch=16, xlab='Permeability',
ylab='Depth, m', xlim=c(0,16000), cex=1.5, cex.lab=1.5, cex.axis=1.2)
grid()
par(new = TRUE)
plot(y=y<-(data$depth),ylim=rev(range(data$depth)),x=x<-(k.core.pred2),type="l", col="darkgreen", lwd = 5, pch=15, xlab='',
ylab='Depth, m', xlim=c(0,16000), cex=1.5, cex.lab=1.5, cex.axis=1.2, main='AdjR.sq1=0.6367 & RMSE=1346')
grid()
legend('topright', legend=c("Observed", "Predicted"), pch=c(16,15), col=c("red", "darkgreen"))
Core Permeability Modeling by using BOX-COX Transforamtion method :
require(MASS)
## Loading required package: MASS
library(MASS)
model1 <- lm(k.core ~ .-1,data=data) #k.core as function of all data
bc <- boxcox(model1, lambda=seq(0,1,by=0.1))
lambda <- bc$x[which.max(bc$y)]
lambda
## [1] 0.3333333
k.corebc <- ((y^lambda-1)/lambda)
model3 <- lm(k.corebc ~ .-1,data=data)
summary(model3)
##
## Call:
## lm(formula = k.corebc ~ . - 1, data = data)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.0037364 -0.0004842 0.0000603 0.0005479 0.0022961
##
## Coefficients: (1 not defined because of singularities)
## Estimate Std. Error t value Pr(>|t|)
## depth 3.051e-03 1.123e-06 2716.209 < 2e-16 ***
## caliper -1.526e-02 6.329e-04 -24.117 < 2e-16 ***
## ind.deep 4.249e-06 1.460e-06 2.909 0.003723 **
## ind.med -5.410e-06 1.604e-06 -3.373 0.000779 ***
## gamma 1.996e-05 3.862e-06 5.169 2.98e-07 ***
## phi.N -5.146e-04 9.167e-04 -0.561 0.574694
## R.deep 1.490e-05 3.935e-06 3.786 0.000164 ***
## R.med -1.955e-05 5.758e-06 -3.395 0.000719 ***
## SP 8.346e-06 1.957e-06 4.264 2.24e-05 ***
## density.corr 1.271e-02 2.984e-03 4.260 2.29e-05 ***
## density 3.184e-03 7.296e-04 4.364 1.44e-05 ***
## phi.core -4.254e-06 1.477e-05 -0.288 0.773494
## k.core 1.675e-08 2.195e-08 0.763 0.445631
## FaciesF1 3.333e+01 1.102e-02 3023.454 < 2e-16 ***
## FaciesF10 3.333e+01 1.105e-02 3016.997 < 2e-16 ***
## FaciesF2 3.333e+01 1.102e-02 3023.717 < 2e-16 ***
## FaciesF3 3.333e+01 1.106e-02 3013.784 < 2e-16 ***
## FaciesF5 3.333e+01 1.106e-02 3014.071 < 2e-16 ***
## FaciesF7 3.333e+01 1.104e-02 3018.238 < 2e-16 ***
## FaciesF8 3.333e+01 1.108e-02 3007.587 < 2e-16 ***
## FaciesF9 3.333e+01 1.115e-02 2990.587 < 2e-16 ***
## phi.core.frac NA NA NA NA
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.0007827 on 798 degrees of freedom
## Multiple R-squared: 1, Adjusted R-squared: 1
## F-statistic: 1.664e+11 on 21 and 798 DF, p-value: < 2.2e-16
par(mfrow=c(2,2))
plot(model3) #plot of model3 before reduction
model3.red <- lm(k.corebc ~ caliper+gamma+phi.N+density+phi.core+Facies-1,data=data)
summary(model3.red)
##
## Call:
## lm(formula = k.corebc ~ caliper + gamma + phi.N + density + phi.core +
## Facies - 1, data = data)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.43160 -0.06697 -0.00998 0.05328 0.61040
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## caliper -1.1912285 0.0638615 -18.653 < 2e-16 ***
## gamma -0.0049245 0.0004677 -10.529 < 2e-16 ***
## phi.N 1.0145357 0.1259097 8.058 2.80e-15 ***
## density 0.7614501 0.1005018 7.576 9.75e-14 ***
## phi.core 0.0021998 0.0020781 1.059 0.29
## FaciesF1 59.5804034 0.6236074 95.542 < 2e-16 ***
## FaciesF10 59.6446831 0.6272407 95.091 < 2e-16 ***
## FaciesF2 59.5098747 0.6284071 94.700 < 2e-16 ***
## FaciesF3 59.6545207 0.6296360 94.744 < 2e-16 ***
## FaciesF5 59.7201098 0.6157497 96.988 < 2e-16 ***
## FaciesF7 59.6547122 0.6157867 96.876 < 2e-16 ***
## FaciesF8 59.7670655 0.6135175 97.417 < 2e-16 ***
## FaciesF9 60.1230777 0.6166642 97.497 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.1172 on 806 degrees of freedom
## Multiple R-squared: 1, Adjusted R-squared: 1
## F-statistic: 1.198e+07 on 13 and 806 DF, p-value: < 2.2e-16
par(mfrow=c(2,2))
plot(model3.red) #plot of model3 after reduction
k.core.pred3 <- predict(model3.red,data) #prediction of k.core after reduction data of model3
k.core.pred3 <-(1+(lambda*k.corebc))^(1/lambda)
AdjR.sq3 <- 1-sum((k.core.pred3 - data$k.core)^2)/sum((data$k.core-mean(data$k.core))^2)
AdjR.sq3
## [1] -2.62226
mspe.model3 <- sqrt(sum((k.core.pred3 - data$k.core)^2)/nrow(data))
mspe.model3
## [1] 4252.266
par(mfrow=c(1,1))
plot(y=y<-(data$depth),ylim=rev(range(data$depth)),x=x<-(data$k.core),type="p", col="red", lwd = 5, pch=16, xlab='Permeability',
ylab='Depth, m', xlim=c(0,16000), cex=1.5, cex.lab=1.5, cex.axis=1.2)
grid()
par(new = TRUE)
plot(y=y<-(data$depth),ylim=rev(range(data$depth)),x=x<-(k.core.pred2),type="l", col="gold", lwd = 5, pch=15, xlab='',
ylab='Depth, m', xlim=c(0,16000), cex=1.5, cex.lab=1.5, cex.axis=1.2, main='AdjR.sq1=0.9999 & RMSE=0')
grid()
legend('topright', legend=c("Observed", "Predicted"), pch=c(16,15), col=c("red", "gold"))
randomly ordered
set.seed(42)
rows <- sample(nrow(data))
data<-data[rows, ]
head(data)
## depth caliper ind.deep ind.med gamma phi.N R.deep R.med SP
## 561 5950.0 8.489 26.833 29.266 41.036 0.284 37.268 34.169 -21.363
## 321 5827.0 8.588 8.425 10.695 24.896 0.191 118.698 93.501 -50.852
## 153 5743.0 8.816 341.817 353.069 72.168 0.248 2.925 2.832 -35.805
## 74 5703.5 8.686 32.016 36.599 56.703 0.080 31.234 27.323 -35.935
## 228 5780.5 8.686 222.486 199.552 60.909 0.177 4.495 5.011 -44.079
## 146 5739.5 8.816 316.687 306.366 68.606 0.268 3.158 3.264 -39.945
## density.corr density phi.core k.core Facies phi.core.frac
## 561 -0.004 2.054 31.2625 3293.4705 F8 0.312625
## 321 -0.021 2.073 27.6898 5338.7207 F5 0.276898
## 153 0.016 2.171 23.5965 467.5622 F3 0.235965
## 74 -0.029 1.943 31.0000 1866.9200 F1 0.310000
## 228 -0.015 2.181 23.1149 1109.5569 F3 0.231149
## 146 -0.038 2.113 29.0670 597.4316 F3 0.290670
split <- round(nrow(data)*0.70)
train <- data[1:split, ]
test <- data[(split+1):nrow(data), ]
mod <- lm(k.core ~.-1,data=train)
summary(mod)
##
## Call:
## lm(formula = k.core ~ . - 1, data = train)
##
## Residuals:
## Min 1Q Median 3Q Max
## -5414.4 -522.6 48.7 509.1 9856.7
##
## Coefficients: (1 not defined because of singularities)
## Estimate Std. Error t value Pr(>|t|)
## depth 6.674 2.126 3.140 0.00178 **
## caliper 241.030 1186.630 0.203 0.83911
## ind.deep 1.607 2.706 0.594 0.55284
## ind.med -1.257 2.971 -0.423 0.67247
## gamma -44.580 7.171 -6.216 1.00e-09 ***
## phi.N -1917.013 1807.525 -1.061 0.28935
## R.deep -17.012 7.109 -2.393 0.01704 *
## R.med 47.590 10.493 4.535 7.05e-06 ***
## SP -3.465 3.671 -0.944 0.34554
## density.corr -2890.898 5709.730 -0.506 0.61284
## density 2493.847 1439.597 1.732 0.08377 .
## phi.core 215.722 27.953 7.717 5.58e-14 ***
## FaciesF1 -47962.532 20633.714 -2.324 0.02046 *
## FaciesF10 -46434.471 20683.769 -2.245 0.02516 *
## FaciesF2 -46513.129 20621.490 -2.256 0.02449 *
## FaciesF3 -47009.538 20703.937 -2.271 0.02356 *
## FaciesF5 -47105.613 20705.530 -2.275 0.02329 *
## FaciesF7 -47150.391 20678.517 -2.280 0.02298 *
## FaciesF8 -48836.257 20737.279 -2.355 0.01887 *
## FaciesF9 -50192.276 20858.403 -2.406 0.01644 *
## phi.core.frac NA NA NA NA
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1244 on 553 degrees of freedom
## Multiple R-squared: 0.8335, Adjusted R-squared: 0.8274
## F-statistic: 138.4 on 20 and 553 DF, p-value: < 2.2e-16
mod_red<-lm(k.core ~ depth+caliper+gamma+R.deep+R.med+phi.core+Facies-1,data=train)
summary(mod_red)
##
## Call:
## lm(formula = k.core ~ depth + caliper + gamma + R.deep + R.med +
## phi.core + Facies - 1, data = train)
##
## Residuals:
## Min 1Q Median 3Q Max
## -5310.2 -515.2 41.9 489.9 9742.9
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## depth 7.181 1.530 4.693 3.39e-06 ***
## caliper 591.301 1032.331 0.573 0.56702
## gamma -45.579 5.208 -8.752 < 2e-16 ***
## R.deep -14.659 6.831 -2.146 0.03231 *
## R.med 42.671 9.940 4.293 2.08e-05 ***
## phi.core 202.739 23.841 8.504 < 2e-16 ***
## FaciesF1 -48477.805 16112.627 -3.009 0.00274 **
## FaciesF10 -46771.173 16229.774 -2.882 0.00411 **
## FaciesF2 -47082.589 16181.588 -2.910 0.00376 **
## FaciesF3 -47390.015 16223.461 -2.921 0.00363 **
## FaciesF5 -47571.622 16230.171 -2.931 0.00352 **
## FaciesF7 -47921.184 16161.698 -2.965 0.00315 **
## FaciesF8 -49553.036 16243.816 -3.051 0.00239 **
## FaciesF9 -50688.624 16362.282 -3.098 0.00205 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1243 on 559 degrees of freedom
## Multiple R-squared: 0.8319, Adjusted R-squared: 0.8277
## F-statistic: 197.7 on 14 and 559 DF, p-value: < 2.2e-16
bc2 <- boxcox(mod_red)
lambda2 <- bc$x[which.max(bc$y)]
kcox2 <- ((train$k.core^lambda-1)/lambda)
par(mfrow=c(1,2))
hist(kcox2)
hist(data$k.core)
model4<-lm(kcox2~.-1,data=train)
summary(model4)
##
## Call:
## lm(formula = kcox2 ~ . - 1, data = train)
##
## Residuals:
## Min 1Q Median 3Q Max
## -18.144 -1.529 0.271 2.036 8.665
##
## Coefficients: (1 not defined because of singularities)
## Estimate Std. Error t value Pr(>|t|)
## depth 2.843e-05 6.468e-03 0.004 0.99649
## caliper -8.725e+00 3.579e+00 -2.438 0.01509 *
## ind.deep -5.958e-03 8.164e-03 -0.730 0.46582
## ind.med 2.390e-03 8.963e-03 0.267 0.78982
## gamma -4.466e-02 2.237e-02 -1.996 0.04640 *
## phi.N -1.799e+01 5.457e+00 -3.297 0.00104 **
## R.deep 1.065e-02 2.155e-02 0.494 0.62139
## R.med -4.661e-02 3.223e-02 -1.446 0.14870
## SP -8.704e-03 1.108e-02 -0.786 0.43246
## density.corr 3.772e+00 1.723e+01 0.219 0.82673
## density 9.264e+00 4.354e+00 2.128 0.03380 *
## phi.core 9.092e-01 8.873e-02 10.246 < 2e-16 ***
## k.core 4.205e-03 1.283e-04 32.785 < 2e-16 ***
## FaciesF1 6.120e+01 6.254e+01 0.979 0.32823
## FaciesF10 6.280e+01 6.267e+01 1.002 0.31670
## FaciesF2 6.109e+01 6.248e+01 0.978 0.32866
## FaciesF3 6.097e+01 6.274e+01 0.972 0.33152
## FaciesF5 6.159e+01 6.274e+01 0.982 0.32671
## FaciesF7 6.503e+01 6.266e+01 1.038 0.29980
## FaciesF8 6.213e+01 6.286e+01 0.988 0.32339
## FaciesF9 6.056e+01 6.324e+01 0.958 0.33868
## phi.core.frac NA NA NA NA
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 3.753 on 552 degrees of freedom
## Multiple R-squared: 0.9881, Adjusted R-squared: 0.9877
## F-statistic: 2190 on 21 and 552 DF, p-value: < 2.2e-16
model4_red<-lm(kcox2~gamma+phi.N+density+phi.core+Facies-1,data=train)
summary(model4_red)
##
## Call:
## lm(formula = kcox2 ~ gamma + phi.N + density + phi.core + Facies -
## 1, data = train)
##
## Residuals:
## Min 1Q Median 3Q Max
## -25.9117 -3.4276 0.3054 4.2740 22.5922
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## gamma -0.27318 0.03166 -8.629 < 2e-16 ***
## phi.N -19.95498 8.64226 -2.309 0.02131 *
## density 19.08967 7.11184 2.684 0.00749 **
## phi.core 2.03220 0.14121 14.392 < 2e-16 ***
## FaciesF1 -46.90754 15.91097 -2.948 0.00333 **
## FaciesF10 -38.07963 16.76794 -2.271 0.02353 *
## FaciesF2 -42.00006 16.19334 -2.594 0.00974 **
## FaciesF3 -43.89477 16.60413 -2.644 0.00843 **
## FaciesF5 -37.13203 16.33217 -2.274 0.02337 *
## FaciesF7 -37.91485 16.44022 -2.306 0.02146 *
## FaciesF8 -44.34360 16.13491 -2.748 0.00618 **
## FaciesF9 -51.45684 16.36175 -3.145 0.00175 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 6.666 on 561 degrees of freedom
## Multiple R-squared: 0.962, Adjusted R-squared: 0.9612
## F-statistic: 1183 on 12 and 561 DF, p-value: < 2.2e-16
k4<- predict(model4_red,data= test)
k4<-(k4*lambda2+1)^(1/lambda2)
rmse4<-sqrt(sum((k4 - test$k.core)^2)/nrow(test))
## Warning in k4 - test$k.core: longer object length is not a multiple of shorter
## object length
par(mfrow=c(1,1))
plot(y=y<-(train$depth),ylim=rev(range(train$depth)),x=x<-(train$k.core),type="p", col="red", lwd = 3, pch=20, xlab='Permeability, md',
ylab='Depth, m', xlim=c(0,16000), cex=1.5, cex.lab=1.5, cex.axis=1.2, main='RMSE=4362.062')
grid()
par(new = TRUE)
plot(y=y<-(train$depth),ylim=rev(range(train$depth)),x=x<-(k4),type="p", col="darkblue", lwd = 3, pch=20, xlab='',
ylab='Depth, m', xlim=c(0,16000), cex=1.5, cex.lab=1.5, cex.axis=1.2,)
grid()
legend('topright', legend=c("Observed", "Predicted"), pch=c(16,15), col=c("red", "darkblue"))
xtable
library(xtable)
xtable(summary(model3.red))
## % latex table generated in R 4.1.2 by xtable 1.8-4 package
## % Sun Jan 16 05:34:52 2022
## \begin{table}[ht]
## \centering
## \begin{tabular}{rrrrr}
## \hline
## & Estimate & Std. Error & t value & Pr($>$$|$t$|$) \\
## \hline
## caliper & -1.1912 & 0.0639 & -18.65 & 0.0000 \\
## gamma & -0.0049 & 0.0005 & -10.53 & 0.0000 \\
## phi.N & 1.0145 & 0.1259 & 8.06 & 0.0000 \\
## density & 0.7615 & 0.1005 & 7.58 & 0.0000 \\
## phi.core & 0.0022 & 0.0021 & 1.06 & 0.2901 \\
## FaciesF1 & 59.5804 & 0.6236 & 95.54 & 0.0000 \\
## FaciesF10 & 59.6447 & 0.6272 & 95.09 & 0.0000 \\
## FaciesF2 & 59.5099 & 0.6284 & 94.70 & 0.0000 \\
## FaciesF3 & 59.6545 & 0.6296 & 94.74 & 0.0000 \\
## FaciesF5 & 59.7201 & 0.6157 & 96.99 & 0.0000 \\
## FaciesF7 & 59.6547 & 0.6158 & 96.88 & 0.0000 \\
## FaciesF8 & 59.7671 & 0.6135 & 97.42 & 0.0000 \\
## FaciesF9 & 60.1231 & 0.6167 & 97.50 & 0.0000 \\
## \hline
## \end{tabular}
## \end{table}
karpur2 <- cbind(data,k.core.pred2)
write.csv(karpur2,"karpur2.csv")
head(karpur2)
## depth caliper ind.deep ind.med gamma phi.N R.deep R.med SP
## 561 5950.0 8.489 26.833 29.266 41.036 0.284 37.268 34.169 -21.363
## 321 5827.0 8.588 8.425 10.695 24.896 0.191 118.698 93.501 -50.852
## 153 5743.0 8.816 341.817 353.069 72.168 0.248 2.925 2.832 -35.805
## 74 5703.5 8.686 32.016 36.599 56.703 0.080 31.234 27.323 -35.935
## 228 5780.5 8.686 222.486 199.552 60.909 0.177 4.495 5.011 -44.079
## 146 5739.5 8.816 316.687 306.366 68.606 0.268 3.158 3.264 -39.945
## density.corr density phi.core k.core Facies phi.core.frac k.core.pred2
## 561 -0.004 2.054 31.2625 3293.4705 F8 0.312625 1460.992
## 321 -0.021 2.073 27.6898 5338.7207 F5 0.276898 1216.419
## 153 0.016 2.171 23.5965 467.5622 F3 0.235965 1187.019
## 74 -0.029 1.943 31.0000 1866.9200 F1 0.310000 2054.629
## 228 -0.015 2.181 23.1149 1109.5569 F3 0.231149 2513.148
## 146 -0.038 2.113 29.0670 597.4316 F3 0.290670 3123.951