Matrix CNC Reliability
rm(list = ls()) # remove all environment
options(warnPartialMatchAttr = T)
library(WeibullR)
library(readxl)
library(ggplot2)
library(magrittr)
source('~/CNC Matrix/itais_functions.R')
Blmr <- read_excel("Breton.xlsx", col_types = c("skip", "skip", "skip", "date", "text", "text",
"text", "text", "text", "text", "text",
"text", "text", "text", "numeric", "numeric",
"numeric", "numeric", "numeric", "numeric",
"numeric", "numeric"))
date <- "2012/01/01"
Blmr$Event<-as.numeric(as.Date(Blmr$END.DATE)-as.Date(date))
q <- wblr(Blmr$Event, pch = 0)
q1 <- wblr.conf(wblr.fit(q, dist = "weibull2p"))
plot.wblr(q1, in.legend.blives = F)
q2 <- wblr.conf(wblr.fit(q, dist = "lognormal"))
plot.wblr(q2, in.legend.blives = F)
q3 <- wblr.fit(q, dist = "weibull3p")
plot.wblr(q3)
q4 <- wblr.fit(q, dist = "lognormal3p")
plot.wblr(q4)
b<-Blmr[Blmr$EQUIP.CODE %in% c("CY10194","CY10194-2"),]
competeRisk(x = b, title = "Matrix CNC", pch = 1, date = date)
## bigIx start stop intercept slope adjR2
## 1 2 1 95 -0.4589068 0.0006070961 0.9744379
## 2 2 96 190 -0.8767926 0.0008159014 0.9345281
## 3 3 1 63 -0.5966098 0.0007428648 0.9811105
## 4 3 64 126 -0.2746622 0.0004794919 0.9881893
## 5 3 127 189 -1.7580328 0.0012342302 0.9598429
## 6 4 1 48 -0.5313212 0.0006718223 0.9859958
## 7 4 49 96 -0.2235679 0.0004408928 0.9879048
## 8 4 97 144 -0.4143181 0.0005615067 0.9727379
## 9 4 145 192 -2.4738595 0.0015682491 0.9648944
## 10 5 1 38 -0.4865422 0.0006218673 0.9895080
## 11 5 39 76 -0.4369792 0.0006087816 0.9416567
## 12 5 77 114 -0.3518984 0.0005252148 0.9751463
## 13 5 115 152 -0.7493475 0.0007340210 0.9346676
## 14 5 153 190 -3.0621233 0.0018401064 0.9778319
## [1] 190
## 3P weibull:
## 3P log-normal: Bad log-normal 3p solution
## Weibull 2-paprameters and 3-parameters likelyhood
## Log-normal 2-paprameters and 3-parameters likelyhood
## 3P weibull: 3p optimization did not converge
## 3P log-normal:
## Weibull 2-paprameters and 3-parameters likelyhood
## Log-normal 2-paprameters and 3-parameters likelyhood
## 3P weibull:
## 3P log-normal:
## Weibull 2-paprameters and 3-parameters likelyhood
## Log-normal 2-paprameters and 3-parameters likelyhood
## [[1]]
## Eta Beta Rsqr AbPval
## 1079.394614 9.787809 0.955575 11.149141
##
## [[2]]
## Eta Beta Rsqr AbPval
## 1689.8855598 10.0566301 0.9385645 5.1288939
##
## [[3]]
## Eta Beta Rsqr AbPval
## 2128.5279176 34.6796078 0.9312035 3.8430357
##
## [[4]]
## [1] 0
f2(b ,"Matrix CNC",MTTF = (q1$fit[[1]]$eta * gamma(1 / q1$fit[[1]]$beta + 1)))
## Matrix CNC Availability:
## Weibull MTTF 1367.00513070555
## Events MTBF: 60.1481481481481
## Events MTTR: 4.5231981981982
## Events MLDT: 46.8108108108108
## Technical: 0.93
## Operational: 0.54
qe <- wblr(Blmr$Event[Blmr$EQUIP.CODE=="CY10192"], pch = 0)
q1 <- wblr.conf(wblr.fit(qe, dist = "weibull2p"))
plot.wblr(q1, in.legend.blives = F)
q2 <- wblr.conf(wblr.fit(qe, dist = "lognormal"))
plot.wblr(q2, in.legend.blives = F)
q3 <- wblr.fit(qe, dist = "weibull3p")
plot.wblr(q3)
q4 <- wblr.fit(qe, dist = "lognormal3p")
plot.wblr(q4)
b<-Blmr[Blmr$EQUIP.CODE %in% c("CY10194"),]
competeRisk(x = b, title = "Eagle CNC", pch = 2, date = date)
## bigIx start stop intercept slope adjR2
## 1 2 1 85 -0.5663888 0.0007299519 0.9776601
## 2 2 86 170 -0.6035634 0.0007070621 0.9814741
## 3 3 1 57 -0.6499882 0.0008122536 0.9777884
## 4 3 58 114 -0.2839845 0.0005208304 0.9865778
## 5 3 115 171 -0.8837958 0.0008423793 0.9703400
## 6 4 1 42 -0.5625952 0.0007161513 0.9884198
## 7 4 43 84 -0.3539343 0.0005711615 0.9625116
## 8 4 85 126 -0.3973266 0.0005873243 0.9760065
## 9 4 127 168 -0.9846827 0.0008899640 0.9697023
## 10 4 169 210 -1.5489949 0.0011567845 NaN
## 11 5 1 34 -0.5356348 0.0006857463 0.9866836
## 12 5 35 68 -0.7006031 0.0008604840 0.9349616
## 13 5 69 102 -0.2246284 0.0004780613 0.9767320
## 14 5 103 136 -0.3607409 0.0005705126 0.9561409
## 15 5 137 170 -1.0940403 0.0009419335 0.9425322
## [1] 170
## 3P weibull:
## 3P log-normal:
## Weibull 2-paprameters and 3-parameters likelyhood
## Log-normal 2-paprameters and 3-parameters likelyhood
## 3P weibull:
## 3P log-normal: Bad log-normal 3p solution
## Weibull 2-paprameters and 3-parameters likelyhood
## Log-normal 2-paprameters and 3-parameters likelyhood
## [[1]]
## Eta Beta Rsqr AbPval
## 1191.4140912 7.2823776 0.9363845 2.8917027
##
## [[2]]
## Eta Beta Rsqr AbPval
## 1995.3882920 11.9281006 0.9223948 1.3523012
f2(b,"Eagle CNC",MTTF = (q1$fit[[1]]$eta * gamma(1 / q1$fit[[1]]$beta + 1)))
## Eagle CNC Availability:
## Weibull MTTF 1198.67851199963
## Events MTBF: 67.1715976331361
## Events MTTR: 4.42067901234568
## Events MLDT: 43.2592592592593
## Technical: 0.938
## Operational: 0.585
a<-q$data$dpoints
ae<-qe$data$dpoints
ggplot(data = a,mapping = aes(time,ppp)) + geom_point(color="black") + geom_smooth(color="red")
## `geom_smooth()` using method = 'loess'
ggplot(data = a,mapping = aes(time,ppp)) + stat_density2d()
a<-Blmr$DownTime[is.numeric(Blmr$DownTime) & !is.na(Blmr$DownTime) & Blmr$DownTime!=0 & Blmr$LDT==0]
plot(a)
hist(a)
summary(a)
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.9167 1.5000 2.0000 2.4038 3.0000 9.5000
a1<-wblr(a,dist="weibull3p")
plot.wblr(wblr.fit(a1),main="Repair Distribution")
a2<-a1$data$dpoints
ggplot(a2,aes(time,ppp))+ geom_point(color="black") + geom_smooth(color="red")
## `geom_smooth()` using method = 'loess'
a3<-a2$time
write.dcf(a3, file = "repair time")