HW4 Sols
Jacob LaRiviere
April 29, 2020
R Markdown
This is an R Markdown document. For more details on using R Markdown see http://rmarkdown.rstudio.com.
When you click the Knit button a document will be generated that includes both content as well as the output of any embedded R code chunks within the document.
The syntax on line 19 then the subsequent on line 56 tells R Markdown to run R code without the asterisks embeds the subsequent code as embedded inside the R Markdown file
library(ggplot2)
library(plyr)
library(dplyr)##
## Attaching package: 'dplyr'## The following objects are masked from 'package:plyr':
##
## arrange, count, desc, failwith, id, mutate, rename, summarise,
## summarize## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionlibrary(reshape2)
setwd("C:/Users/jlariv/OneDrive/Econ 404/")
oj <- read.csv("oj.csv")
ggplot(oj, aes(logmove, price)) + geom_point(aes(color = factor(brand)))
#First why to add in lagged weeks
df1 <-oj
df1$week<-df1$week+1 # df1 now has NEXT week and not the current one. If we merge this by weeks now, this is last week's price (e.g., "lagged price").
myvars <- c("price", "week", "brand","store")
df1 <- df1[myvars]
lagged <- merge(oj, df1, by=c("brand","store","week"))
#NOTE: The number of observations decreased. Why? You've just lost (at least) one week's worth of data at each store.
lagged=lagged[order(lagged$week,lagged$store),]
lagged=lagged[order(lagged$store,lagged$week),]
#Comparing this to above, Store two is nowhere to be found. There was missing data for consecutive weeks. As a result, it gets dropped.
colnames(lagged)[18] <- "lagged_price"
colnames(lagged)[6] <- "price"
#Explicit code to be super celar about what cross validation is doing.
set.seed(9)
folds<-5
random_lagged <- lagged[sample(nrow(lagged)),]
random_lagged$rand_obs<-seq(1,nrow(random_lagged))
# %% is the modulus operator in R
random_lagged$partition <- random_lagged$rand_obs %% folds +1
MSEs <- c(1:folds)
#For people who've never seen a for loop, in this case its some that helps you iterate through code. "i" is a count variable that runs from 1 to 5 (e.g., folds <- 5 in like 47) and takes the value of the iteration round of the loop within the loop (e.g., within the brackets)
for (i in 1:folds) {
oj_test1 <- random_lagged[which(random_lagged$partition==i),]
oj_train1 <- anti_join(random_lagged,oj_test1)
reg1 <- lm(logmove ~ log(price) + feat + brand + brand*log(price) + AGE60 + EDUC + ETHNIC + INCOME + HHLARGE + WORKWOM + HVAL150 + SSTRDIST + SSTRVOL + CPDIST5 + CPWVOL5 + EDUC*log(price) + HHLARGE*log(price) + log(lagged_price) , data=oj_train1)
# Predict y
oj_test1$logmove_hat <- predict(reg1, newdata=oj_test1)
MSE <- mean((oj_test1$logmove_hat - oj_test1$logmove)^2)
MSEs[i] <- MSE
}## Joining, by = c("brand", "store", "week", "logmove", "feat", "price", "AGE60", "EDUC", "ETHNIC", "INCOME", "HHLARGE", "WORKWOM", "HVAL150", "SSTRDIST", "SSTRVOL", "CPDIST5", "CPWVOL5", "lagged_price", "rand_obs", "partition")## Joining, by = c("brand", "store", "week", "logmove", "feat", "price", "AGE60", "EDUC", "ETHNIC", "INCOME", "HHLARGE", "WORKWOM", "HVAL150", "SSTRDIST", "SSTRVOL", "CPDIST5", "CPWVOL5", "lagged_price", "rand_obs", "partition")
## Joining, by = c("brand", "store", "week", "logmove", "feat", "price", "AGE60", "EDUC", "ETHNIC", "INCOME", "HHLARGE", "WORKWOM", "HVAL150", "SSTRDIST", "SSTRVOL", "CPDIST5", "CPWVOL5", "lagged_price", "rand_obs", "partition")
## Joining, by = c("brand", "store", "week", "logmove", "feat", "price", "AGE60", "EDUC", "ETHNIC", "INCOME", "HHLARGE", "WORKWOM", "HVAL150", "SSTRDIST", "SSTRVOL", "CPDIST5", "CPWVOL5", "lagged_price", "rand_obs", "partition")
## Joining, by = c("brand", "store", "week", "logmove", "feat", "price", "AGE60", "EDUC", "ETHNIC", "INCOME", "HHLARGE", "WORKWOM", "HVAL150", "SSTRDIST", "SSTRVOL", "CPDIST5", "CPWVOL5", "lagged_price", "rand_obs", "partition")# Printout of vector of MSEs and average of MSEs
MSEs## [1] 0.4177313 0.4113887 0.4107742 0.4152023 0.4191698mean(MSEs)## [1] 0.4148533for (i in 1:folds) {
oj_test1 <- random_lagged[which(random_lagged$partition==i),]
oj_train1 <- anti_join(random_lagged,oj_test1)
reg1 <- lm(logmove ~ log(price) + feat+brand + AGE60 + EDUC + ETHNIC + INCOME + HHLARGE + WORKWOM + HVAL150 + SSTRDIST + SSTRVOL + CPDIST5 + CPWVOL5, data=oj_train1)
# Predict y
oj_test1$logmove_hat <- predict(reg1, newdata=oj_test1)
MSE <- mean((oj_test1$logmove_hat - oj_test1$logmove)^2)
MSEs[i] <- MSE
}## Joining, by = c("brand", "store", "week", "logmove", "feat", "price", "AGE60", "EDUC", "ETHNIC", "INCOME", "HHLARGE", "WORKWOM", "HVAL150", "SSTRDIST", "SSTRVOL", "CPDIST5", "CPWVOL5", "lagged_price", "rand_obs", "partition")
## Joining, by = c("brand", "store", "week", "logmove", "feat", "price", "AGE60", "EDUC", "ETHNIC", "INCOME", "HHLARGE", "WORKWOM", "HVAL150", "SSTRDIST", "SSTRVOL", "CPDIST5", "CPWVOL5", "lagged_price", "rand_obs", "partition")
## Joining, by = c("brand", "store", "week", "logmove", "feat", "price", "AGE60", "EDUC", "ETHNIC", "INCOME", "HHLARGE", "WORKWOM", "HVAL150", "SSTRDIST", "SSTRVOL", "CPDIST5", "CPWVOL5", "lagged_price", "rand_obs", "partition")
## Joining, by = c("brand", "store", "week", "logmove", "feat", "price", "AGE60", "EDUC", "ETHNIC", "INCOME", "HHLARGE", "WORKWOM", "HVAL150", "SSTRDIST", "SSTRVOL", "CPDIST5", "CPWVOL5", "lagged_price", "rand_obs", "partition")
## Joining, by = c("brand", "store", "week", "logmove", "feat", "price", "AGE60", "EDUC", "ETHNIC", "INCOME", "HHLARGE", "WORKWOM", "HVAL150", "SSTRDIST", "SSTRVOL", "CPDIST5", "CPWVOL5", "lagged_price", "rand_obs", "partition")MSEs## [1] 0.4537844 0.4457258 0.4452941 0.4500627 0.4509782mean(MSEs)## [1] 0.449169#We see that this simpler model has a HIGHER out of sample MSE. Not as good at prediction.
library(reshape2)
library(glmnet)## Loading required package: Matrix## Loaded glmnet 3.0-2###################
# In Class
###################
x <- model.matrix(~ log(price) + feat + brand + brand*log(price) + AGE60 + EDUC + ETHNIC + INCOME + HHLARGE + WORKWOM + HVAL150 + SSTRDIST + SSTRVOL + CPDIST5 + CPWVOL5 + EDUC*log(price) + HHLARGE*log(price) + log(lagged_price) , data=lagged)
y <- as.numeric(as.matrix(lagged$logmove))
set.seed(720)
#lasso_v1 <- cv.glmnet(x, y, alpha=1)
lasso_v1 <- glmnet(x, y, alpha=1)
plot(lasso_v1)
coef(lasso_v1, s=lasso_v1$lambda.min)## 22 x 73 sparse Matrix of class "dgCMatrix"## [[ suppressing 73 column names 's0', 's1', 's2' ... ]]##
## (Intercept) 9.174759 9.1475029 9.1226678 9.13710438 9.1989982
## (Intercept) . . . . .
## log(price) . . . -0.04473513 -0.1443215
## feat . 0.1151777 0.2201233 0.30708480 0.3749318
## brandminute.maid . . . . .
## brandtropicana . . . . .
## AGE60 . . . . .
## EDUC . . . . .
## ETHNIC . . . . .
## INCOME . . . . .
## HHLARGE . . . . .
## WORKWOM . . . . .
## HVAL150 . . . . .
## SSTRDIST . . . . .
## SSTRVOL . . . . .
## CPDIST5 . . . . .
## CPWVOL5 . . . . .
## log(lagged_price) . . . . .
## log(price):brandminute.maid . . . . .
## log(price):brandtropicana . . . . .
## log(price):EDUC . . . . .
## log(price):HHLARGE . . . . .
##
## (Intercept) 9.2553793 9.3067516 9.3535602 9.3968030
## (Intercept) . . . .
## log(price) -0.2350437 -0.3177064 -0.3930256 -0.4487721
## feat 0.4367547 0.4930855 0.5444121 0.5911069
## brandminute.maid . . . .
## brandtropicana . . . .
## AGE60 . . . .
## EDUC . . . .
## ETHNIC . . . .
## INCOME . . . .
## HHLARGE . . . .
## WORKWOM . . . .
## HVAL150 . . . .
## SSTRDIST . . . .
## SSTRVOL . . . .
## CPDIST5 . . . .
## CPWVOL5 . . . .
## log(lagged_price) . . . .
## log(price):brandminute.maid . . . .
## log(price):brandtropicana . . . .
## log(price):EDUC . . . .
## log(price):HHLARGE . . . -0.1181259
##
## (Intercept) 9.4365444 9.4727392 9.5058397 9.5358810
## (Intercept) . . . .
## log(price) -0.4766223 -0.5019428 -0.5254306 -0.5464238
## feat 0.6335646 0.6722511 0.7074957 0.7396143
## brandminute.maid . . . .
## brandtropicana . . . .
## AGE60 . . . .
## EDUC . . . .
## ETHNIC . . . .
## INCOME . . . .
## HHLARGE . . . .
## WORKWOM . . . .
## HVAL150 . . . .
## SSTRDIST . . . .
## SSTRVOL . . . .
## CPDIST5 . . . .
## CPWVOL5 . . . .
## log(lagged_price) . . . .
## log(price):brandminute.maid . . . .
## log(price):brandtropicana . . . .
## log(price):EDUC . . . .
## log(price):HHLARGE -0.4283349 -0.7112909 -0.9668237 -1.2018941
##
## (Intercept) 9.5632509 9.5881892 9.61623575 9.68533198
## (Intercept) . . . .
## log(price) -0.5655427 -0.5829629 -0.60899464 -0.68314422
## feat 0.7688796 0.7955451 0.81906943 0.83659189
## brandminute.maid . . . .
## brandtropicana . . 0.01169695 0.07750111
## AGE60 . . . .
## EDUC . . . .
## ETHNIC . . . .
## INCOME . . . .
## HHLARGE . . . .
## WORKWOM . . . .
## HVAL150 . . . .
## SSTRDIST . . . .
## SSTRVOL . . . .
## CPDIST5 . . . .
## CPWVOL5 . . . -0.03522429
## log(lagged_price) . . . .
## log(price):brandminute.maid . . . .
## log(price):brandtropicana . . . .
## log(price):EDUC . . . .
## log(price):HHLARGE -1.4161324 -1.6113395 -1.80126485 -2.04167300
##
## (Intercept) 9.75904921 9.81279679 9.8557881 9.92731970
## (Intercept) . . . .
## log(price) -0.74335995 -0.82636180 -0.9214419 -1.00126390
## feat 0.85257691 0.85936887 0.8610885 0.86295455
## brandminute.maid . . . .
## brandtropicana 0.13730999 0.18230482 0.2185578 0.25178928
## AGE60 . . . .
## EDUC . . . .
## ETHNIC . . . .
## INCOME . . . .
## HHLARGE . . . .
## WORKWOM . . . -0.09882442
## HVAL150 . . . .
## SSTRDIST . . . .
## SSTRVOL . . . .
## CPDIST5 . . . .
## CPWVOL5 -0.09312002 -0.14554430 -0.1926849 -0.22795106
## log(lagged_price) . 0.05131067 0.1255812 0.19172946
## log(price):brandminute.maid . . . .
## log(price):brandtropicana . . . .
## log(price):EDUC . . . .
## log(price):HHLARGE -2.31780515 -2.56819605 -2.7740235 -3.01158069
##
## (Intercept) 10.0073870 10.09351142 10.17710684 10.2525629092
## (Intercept) . . . .
## log(price) -1.0704042 -1.15500118 -1.24375189 -1.3228261663
## feat 0.8646842 0.86291622 0.85950836 0.8566470896
## brandminute.maid . 0.03948402 0.09006367 0.1355803956
## brandtropicana 0.2822427 0.34635557 0.41842376 0.4834046537
## AGE60 . . . .
## EDUC . . . .
## ETHNIC . . . 0.0006533839
## INCOME . . . .
## HHLARGE . . . .
## WORKWOM -0.2346713 -0.36370906 -0.48279376 -0.5907329848
## HVAL150 . . . .
## SSTRDIST . . . .
## SSTRVOL . . . .
## CPDIST5 . . . .
## CPWVOL5 -0.2567441 -0.28340876 -0.30782016 -0.3299203401
## log(lagged_price) 0.2516464 0.28501888 0.30957612 0.3315362221
## log(price):brandminute.maid . . . .
## log(price):brandtropicana . . . .
## log(price):EDUC . . . .
## log(price):HHLARGE -3.2557762 -3.50103604 -3.72234343 -3.9281089640
##
## (Intercept) 10.31138429 10.374272499 10.43294319 10.48650685
## (Intercept) . . . .
## log(price) -1.39301890 -1.455968230 -1.51347365 -1.56582170
## feat 0.85379926 0.851157168 0.84903703 0.84696981
## brandminute.maid 0.17768538 0.216047975 0.25043141 0.28201053
## brandtropicana 0.54367201 0.598460313 0.64760132 0.69270490
## AGE60 . . . .
## EDUC . . . .
## ETHNIC 0.02632482 0.060759567 0.09407708 0.12442291
## INCOME . . . .
## HHLARGE . . . .
## WORKWOM -0.67766804 -0.772134649 -0.85982988 -0.94015413
## HVAL150 . . . .
## SSTRDIST . . . .
## SSTRVOL . -0.009913619 -0.02083632 -0.03071054
## CPDIST5 . . . .
## CPWVOL5 -0.34359038 -0.342890420 -0.33945802 -0.33654758
## log(lagged_price) 0.35117072 0.369527661 0.38589101 0.40101820
## log(price):brandminute.maid . . . .
## log(price):brandtropicana . . . .
## log(price):EDUC . . . .
## log(price):HHLARGE -4.14422886 -4.346203432 -4.52197633 -4.68569029
##
## (Intercept) 10.5353123 10.5842085195 10.61791750 10.62777250
## (Intercept) . . . .
## log(price) -1.6134864 -1.6621345444 -1.75572571 -1.85180749
## feat 0.8450895 0.8434152080 0.84166411 0.83986805
## brandminute.maid 0.3107810 0.3370975670 0.36188699 0.38490888
## brandtropicana 0.7337989 0.7715316721 0.80754657 0.84093275
## AGE60 . . 0.07554039 0.18546955
## EDUC . . . .
## ETHNIC 0.1520772 0.1787484557 0.21345749 0.24731861
## INCOME . . . .
## HHLARGE . . . .
## WORKWOM -1.0133586 -1.0948078169 -1.17482658 -1.20942250
## HVAL150 . 0.0009283278 . .
## SSTRDIST . . . .
## SSTRVOL -0.0397081 -0.0475039933 -0.05230015 -0.05586669
## CPDIST5 . . . .
## CPWVOL5 -0.3338955 -0.3319953659 -0.33495796 -0.33921115
## log(lagged_price) 0.4147921 0.4268177796 0.43617898 0.44479752
## log(price):brandminute.maid . . . .
## log(price):brandtropicana . . . .
## log(price):EDUC . 0.0193012553 0.15409097 0.28308794
## log(price):HHLARGE -4.8349904 -4.9598911573 -4.89151018 -4.76198773
##
## (Intercept) 10.6392020781 10.65486840 10.67123908 10.668200679
## (Intercept) . . . .
## log(price) -1.9373519842 -2.04227916 -2.13062179 -2.216868439
## feat 0.8382831118 0.83670585 0.83586452 0.835042635
## brandminute.maid 0.4057702199 0.43050895 0.45179589 0.470969985
## brandtropicana 0.8712155426 0.85160644 0.82938439 0.810297848
## AGE60 0.2807911087 0.38024787 0.47616538 0.588598889
## EDUC . . . .
## ETHNIC 0.2779715798 0.30670592 0.34493307 0.385174034
## INCOME . . . .
## HHLARGE . . . .
## WORKWOM -1.2453908398 -1.26950393 -1.28390568 -1.270124189
## HVAL150 . . . .
## SSTRDIST . . -0.00163307 -0.003361582
## SSTRVOL -0.0592425556 -0.06205185 -0.06247905 -0.061891569
## CPDIST5 . . . 0.001980196
## CPWVOL5 -0.3428422754 -0.34674225 -0.35957158 -0.373960160
## log(lagged_price) 0.4525850506 0.46013027 0.46727941 0.474366146
## log(price):brandminute.maid . . . .
## log(price):brandtropicana 0.0003583555 0.05645411 0.10974044 0.156768399
## log(price):EDUC 0.3984581049 0.51088984 0.61375951 0.716297012
## log(price):HHLARGE -4.6564147670 -4.52265020 -4.43169854 -4.317060818
##
## (Intercept) 10.662373400 10.657418048 10.652890227 10.648756787
## (Intercept) . . . .
## log(price) -2.291144573 -2.359313379 -2.421323717 -2.477872562
## feat 0.834368599 0.833764902 0.833211406 0.832714224
## brandminute.maid 0.487978801 0.503657568 0.517923552 0.531066245
## brandtropicana 0.793630286 0.777151182 0.762411921 0.748198734
## AGE60 0.681484496 0.765743887 0.842458010 0.912376440
## EDUC . . . .
## ETHNIC 0.425124236 0.461218165 0.494105534 0.524042138
## INCOME . . . .
## HHLARGE . . . .
## WORKWOM -1.264350852 -1.258868638 -1.253950471 -1.248978508
## HVAL150 . . . .
## SSTRDIST -0.004930196 -0.006352005 -0.007647303 -0.008823047
## SSTRVOL -0.061590974 -0.061322826 -0.061079301 -0.060890733
## CPDIST5 0.006473598 0.010533609 0.014233497 0.017589512
## CPWVOL5 -0.386566213 -0.398026240 -0.408465202 -0.417901992
## log(lagged_price) 0.481208646 0.487358980 0.492968919 0.497980982
## log(price):brandminute.maid . . . .
## log(price):brandtropicana 0.198194226 0.237439631 0.272900983 0.306168963
## log(price):EDUC 0.811820415 0.898556878 0.977573205 1.049132025
## log(price):HHLARGE -4.242496218 -4.173846074 -4.111473932 -4.055487129
##
## (Intercept) 10.698134901 11.01720329 11.26205411 11.48504886
## (Intercept) . . . .
## log(price) -2.529771237 -2.56161332 -2.58540450 -2.60659809
## feat 0.832245117 0.83166938 0.83125168 0.83084903
## brandminute.maid 0.542923254 0.55396118 0.56398792 0.57301129
## brandtropicana 0.735161036 0.72540483 0.71566423 0.70747096
## AGE60 0.980601347 1.10814090 1.21580885 1.31362139
## EDUC . . . .
## ETHNIC 0.546933324 0.54534973 0.54659018 0.54776949
## INCOME -0.005680632 -0.04424726 -0.07463384 -0.10233000
## HHLARGE 0.032096628 0.48015163 0.87224710 1.23212310
## WORKWOM -1.236265691 -1.16270243 -1.09947853 -1.04250735
## HVAL150 . . . .
## SSTRDIST -0.009891778 -0.01098034 -0.01192976 -0.01279971
## SSTRVOL -0.061149328 -0.06239502 -0.06359079 -0.06466157
## CPDIST5 0.020842959 0.02532395 0.02903610 0.03243559
## CPWVOL5 -0.425494405 -0.43016760 -0.43424444 -0.43800789
## log(lagged_price) 0.502759696 0.50738988 0.51131651 0.51503470
## log(price):brandminute.maid . . . .
## log(price):brandtropicana 0.336411621 0.36229022 0.38662196 0.40797923
## log(price):EDUC 1.122132692 1.25396997 1.36093477 1.45867543
## log(price):HHLARGE -4.016568095 -4.25295220 -4.48230648 -4.69445926
##
## (Intercept) 11.68728928 11.87174695 12.02721516 12.06205567
## (Intercept) . . . .
## log(price) -2.62554142 -2.64288623 -2.65904705 -2.69952177
## feat 0.83049200 0.83015727 0.82988132 0.82985528
## brandminute.maid 0.58124157 0.58871758 0.59541746 0.59317741
## brandtropicana 0.69970273 0.69306830 0.68680579 0.68040542
## AGE60 1.40274487 1.48384571 1.55555662 1.59697970
## EDUC . . -0.01071934 -0.12677751
## ETHNIC 0.54888183 0.54988869 0.55034578 0.55885688
## INCOME -0.12750157 -0.15043908 -0.16989712 -0.17358530
## HHLARGE 1.56237184 1.86185436 2.12951617 2.28855135
## WORKWOM -0.99052252 -0.94328160 -0.89817526 -0.86347351
## HVAL150 . . . .
## SSTRDIST -0.01359227 -0.01431390 -0.01492263 -0.01538417
## SSTRVOL -0.06563698 -0.06652630 -0.06737881 -0.06796463
## CPDIST5 0.03552741 0.03834498 0.04066573 0.04166356
## CPWVOL5 -0.44144035 -0.44456296 -0.44725916 -0.44970153
## log(lagged_price) 0.51840757 0.52149557 0.52434456 0.52678222
## log(price):brandminute.maid . . . 0.01202270
## log(price):brandtropicana 0.42774723 0.44529272 0.46132053 0.47851461
## log(price):EDUC 1.54752357 1.62849531 1.70794417 1.86602744
## log(price):HHLARGE -4.89083606 -5.06824569 -5.23411063 -5.35915989
##
## (Intercept) 12.14685868 12.22908341 12.30551063 12.37330408
## (Intercept) . . . .
## log(price) -2.75012303 -2.79773039 -2.84122104 -2.88268536
## feat 0.82996560 0.83001748 0.83003469 0.83006447
## brandminute.maid 0.57941617 0.56958385 0.56127094 0.55337288
## brandtropicana 0.67185388 0.66295560 0.65562883 0.64815059
## AGE60 1.63340822 1.66417822 1.69238213 1.71707634
## EDUC -0.25946511 -0.39372432 -0.51577897 -0.62991502
## ETHNIC 0.56587608 0.57279096 0.57897701 0.58486967
## INCOME -0.18211542 -0.19036232 -0.19801239 -0.20460532
## HHLARGE 2.42098877 2.53496864 2.63530500 2.71888427
## WORKWOM -0.82315491 -0.78469138 -0.74961118 -0.71823970
## HVAL150 0.02355404 0.05278751 0.07937156 0.10364399
## SSTRDIST -0.01586244 -0.01630790 -0.01671330 -0.01708063
## SSTRVOL -0.06702032 -0.06563168 -0.06436671 -0.06321177
## CPDIST5 0.04355582 0.04551341 0.04730215 0.04891286
## CPWVOL5 -0.45377778 -0.45816468 -0.46215646 -0.46578882
## log(lagged_price) 0.52871121 0.53025107 0.53167357 0.53296535
## log(price):brandminute.maid 0.04001360 0.06217480 0.08141326 0.09948929
## log(price):brandtropicana 0.50060607 0.52161438 0.53978181 0.55729630
## log(price):EDUC 2.00534260 2.13282695 2.24911528 2.35755991
## log(price):HHLARGE -5.41111199 -5.44262656 -5.46661454 -5.48121123
##
## (Intercept) 12.43594928 12.49327950 12.54849750 12.59427098
## (Intercept) . . . .
## log(price) -2.92041174 -2.95461803 -2.98440425 -3.01358479
## feat 0.83008678 0.83010993 0.83011783 0.83014102
## brandminute.maid 0.54627138 0.53982713 0.53424091 0.52865658
## brandtropicana 0.64124248 0.63495858 0.63013876 0.62487911
## AGE60 1.73953598 1.76007794 1.78008582 1.79686106
## EDUC -0.73265725 -0.82568029 -0.90778356 -0.98835012
## ETHNIC 0.59012125 0.59487010 0.59891193 0.60311613
## INCOME -0.21069242 -0.21627774 -0.22184321 -0.22624165
## HHLARGE 2.79414288 2.86297663 2.93144584 2.98754837
## WORKWOM -0.68993536 -0.66414588 -0.63971679 -0.61804138
## HVAL150 0.12553650 0.14542717 0.16361450 0.18054600
## SSTRDIST -0.01741497 -0.01771965 -0.01800058 -0.01825366
## SSTRVOL -0.06217196 -0.06122856 -0.06036342 -0.05955436
## CPDIST5 0.05038194 0.05172205 0.05297054 0.05407551
## CPWVOL5 -0.46908949 -0.47209416 -0.47484491 -0.47735778
## log(lagged_price) 0.53415588 0.53523094 0.53620438 0.53709522
## log(price):brandminute.maid 0.11584326 0.13070366 0.14373875 0.15640616
## log(price):brandtropicana 0.57333871 0.58793291 0.60016790 0.61239290
## log(price):EDUC 2.45560073 2.54444957 2.62323644 2.69942441
## log(price):HHLARGE -5.49351754 -5.50496111 -5.51929082 -5.52794173
##
## (Intercept) 12.64125964 12.68028713 12.71553397 12.74863959
## (Intercept) . . . .
## log(price) -3.03849633 -3.06204121 -3.08395395 -3.10334368
## feat 0.83015021 0.83016216 0.83017983 0.83019287
## brandminute.maid 0.52399166 0.51955002 0.51539169 0.51177240
## brandtropicana 0.62090398 0.61681637 0.61277068 0.60915451
## AGE60 1.81374448 1.82792141 1.84058970 1.85223138
## EDUC -1.05612673 -1.12095975 -1.18042300 -1.23209116
## ETHNIC 0.60641350 0.60974757 0.61283126 0.61547820
## INCOME -0.23097914 -0.23478723 -0.23817664 -0.24139706
## HHLARGE 3.04329190 3.09126615 3.13287212 3.17099147
## WORKWOM -0.59753465 -0.57962719 -0.56349783 -0.54910305
## HVAL150 0.19568507 0.20951093 0.22208034 0.23319992
## SSTRDIST -0.01848744 -0.01869739 -0.01888833 -0.01906172
## SSTRVOL -0.05883843 -0.05818197 -0.05758547 -0.05706462
## CPDIST5 0.05511988 0.05604233 0.05687875 0.05764383
## CPWVOL5 -0.47963386 -0.48170706 -0.48359478 -0.48529460
## log(lagged_price) 0.53789145 0.53863662 0.53931073 0.53992654
## log(price):brandminute.maid 0.16726139 0.17747599 0.18697350 0.19538225
## log(price):brandtropicana 0.62252700 0.63232256 0.64162367 0.64997515
## log(price):EDUC 2.76457512 2.82629431 2.88287934 2.93254934
## log(price):HHLARGE -5.53777656 -5.54625786 -5.55201686 -5.55757177
##
## (Intercept) 12.78157400 12.80810907 12.82762813 12.86089275
## (Intercept) . . . .
## log(price) -3.12082467 -3.13720760 -3.14854579 -3.16439162
## feat 0.83020878 0.83021681 0.83023099 0.83026586
## brandminute.maid 0.50841414 0.50533205 0.50348587 0.49986906
## brandtropicana 0.60637117 0.60355343 0.60032582 0.59838138
## AGE60 1.86444289 1.87406691 1.87989866 1.89407436
## EDUC -1.28039657 -1.32540748 -1.34834439 -1.40133742
## ETHNIC 0.61782961 0.62016815 0.62204554 0.62381791
## INCOME -0.24474771 -0.24731531 -0.24911793 -0.25278022
## HHLARGE 3.21063182 3.24277038 3.26143410 3.31101067
## WORKWOM -0.53407898 -0.52183293 -0.51650262 -0.49726451
## HVAL150 0.24410369 0.25364133 0.25871044 0.27102640
## SSTRDIST -0.01922301 -0.01936736 -0.01947186 -0.01962014
## SSTRVOL -0.05656320 -0.05611055 -0.05604008 -0.05533928
## CPDIST5 0.05837388 0.05900672 0.05950327 0.06017110
## CPWVOL5 -0.48684241 -0.48826978 -0.48910950 -0.49067416
## log(lagged_price) 0.54044231 0.54095792 0.54143334 0.54173852
## log(price):brandminute.maid 0.20306913 0.21013829 0.21529590 0.22259818
## log(price):brandtropicana 0.65704637 0.66380377 0.67037988 0.67586509
## log(price):EDUC 2.97842680 3.02120317 3.04583393 3.09313418
## log(price):HHLARGE -5.56416421 -5.56925338 -5.57323017 -5.58479171
##
## (Intercept) 12.87642103 12.89780747 12.90493231 12.92990292
## (Intercept) . . . .
## log(price) -3.17402217 -3.18682042 -3.19406687 -3.20705783
## feat 0.83024707 0.83026862 0.83026343 0.83029950
## brandminute.maid 0.49842194 0.49569086 0.49477851 0.49166543
## brandtropicana 0.59689660 0.59450838 0.59270772 0.59023814
## AGE60 1.89925100 1.90740702 1.90854572 1.91908793
## EDUC -1.42460115 -1.46196823 -1.47676288 -1.51719391
## ETHNIC 0.62553941 0.62713798 0.62873558 0.62997489
## INCOME -0.25424091 -0.25636804 -0.25679775 -0.25945782
## HHLARGE 3.32867054 3.35682261 3.36273272 3.39635380
## WORKWOM -0.49183437 -0.48087490 -0.48027429 -0.46582037
## HVAL150 0.27611748 0.28397038 0.28666848 0.29561875
## SSTRDIST -0.01971012 -0.01982105 -0.01989514 -0.01999645
## SSTRVOL -0.05520987 -0.05477336 -0.05470068 -0.05421182
## CPDIST5 0.06058898 0.06106453 0.06135875 0.06182893
## CPWVOL5 -0.49144710 -0.49260961 -0.49329967 -0.49435346
## log(lagged_price) 0.54217152 0.54250147 0.54289820 0.54309983
## log(price):brandminute.maid 0.22663720 0.23240972 0.23536744 0.24144209
## log(price):brandtropicana 0.68002851 0.68534164 0.68925678 0.69449204
## log(price):EDUC 3.11648337 3.15135602 3.16719347 3.20366536
## log(price):HHLARGE -5.58899801 -5.59406327 -5.59624087 -5.60063467
##
## (Intercept) 12.93034523 12.94801276 12.96802895 12.96787456
## (Intercept) . . . .
## log(price) -3.21226319 -3.21985852 -3.23105939 -3.23286850
## feat 0.83029924 0.83029154 0.83033707 0.83039810
## brandminute.maid 0.49118469 0.48973216 0.48692400 0.48690926
## brandtropicana 0.58872115 0.58698302 0.58488008 0.58456673
## AGE60 1.91680594 1.92398442 1.93268012 1.93145922
## EDUC -1.52591176 -1.54373529 -1.58086944 -1.58361715
## ETHNIC 0.63126794 0.63225496 0.63330449 0.63386590
## INCOME -0.25907449 -0.26097419 -0.26309920 -0.26289539
## HHLARGE 3.39234182 3.41241783 3.44096555 3.43851247
## WORKWOM -0.46877921 -0.46139236 -0.44854951 -0.45037061
## HVAL150 0.29678922 0.30100062 0.30907916 0.30958271
## SSTRDIST -0.02006615 -0.02011196 -0.02020211 -0.02024367
## SSTRVOL -0.05405857 -0.05410951 -0.05358492 -0.05352505
## CPDIST5 0.06202568 0.06234668 0.06273134 0.06285726
## CPWVOL5 -0.49523546 -0.49533130 -0.49637185 -0.49697970
## log(lagged_price) 0.54348487 0.54369181 0.54381876 0.54401387
## log(price):brandminute.maid 0.24341629 0.24693312 0.25215856 0.25279030
## log(price):brandtropicana 0.69756299 0.70136782 0.70567211 0.70663819
## log(price):EDUC 3.21344666 3.23175451 3.26435632 3.26765846
## log(price):HHLARGE -5.60018711 -5.60321941 -5.60671487 -5.60769037# Now ready for cross validation version of the object
cvfit <- cv.glmnet(x, y, alpha=1)
#Results
plot(cvfit)
cvfit$lambda.min## [1] 0.0006793529log(cvfit$lambda.min)## [1] -7.29437coef(cvfit, s = "lambda.min")## 22 x 1 sparse Matrix of class "dgCMatrix"
## 1
## (Intercept) 12.96787456
## (Intercept) .
## log(price) -3.23286850
## feat 0.83039810
## brandminute.maid 0.48690926
## brandtropicana 0.58456673
## AGE60 1.93145922
## EDUC -1.58361715
## ETHNIC 0.63386590
## INCOME -0.26289539
## HHLARGE 3.43851247
## WORKWOM -0.45037061
## HVAL150 0.30958271
## SSTRDIST -0.02024367
## SSTRVOL -0.05352505
## CPDIST5 0.06285726
## CPWVOL5 -0.49697970
## log(lagged_price) 0.54401387
## log(price):brandminute.maid 0.25279030
## log(price):brandtropicana 0.70663819
## log(price):EDUC 3.26765846
## log(price):HHLARGE -5.60769037#LASSO is nice because it is transparent and algorithmic rather than up to the discretion of the econometrician. NOTE: the coefficients of LASSO are different from the coefficients in OLS. They are biased downward.
#dcast is a function in the reshape2 library that turns "long data" into "wide data""
oj_prices <-lagged[,1:6]
oj_wide <- dcast(oj_prices, store + week ~ brand)## Using price as value column: use value.var to override.#New tidyverse package
#oj %>% select(store, week, brand, price) %>% spread(brand, price)
# gather is going wide to long !!!!gather("variable","value", -store, -week) or gather("variable","value", 3:5)
colnames(oj_wide)[3] <- "P_Dom"
colnames(oj_wide)[4] <- "P_MM"
colnames(oj_wide)[5] <- "P_Trop"
oj_cross <- merge(oj, oj_wide, by=c("week","store"))
#Merge the wide data back in then only look at the cross price elasticity matrix for tropicana.
trop_cross <- subset(oj_cross, brand=="tropicana")
regcrosst = glm(logmove ~ log(P_Dom)*feat+log(P_MM)*feat+log(P_Trop)*feat, data=trop_cross)
summary(regcrosst)##
## Call:
## glm(formula = logmove ~ log(P_Dom) * feat + log(P_MM) * feat +
## log(P_Trop) * feat, data = trop_cross)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -3.3152 -0.3802 0.0019 0.3660 2.6480
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.95288 0.04430 247.244 < 2e-16 ***
## log(P_Dom) 0.15027 0.02958 5.081 3.83e-07 ***
## feat 1.48611 0.10089 14.731 < 2e-16 ***
## log(P_MM) 0.27827 0.03813 7.298 3.17e-13 ***
## log(P_Trop) -2.16685 0.03782 -57.288 < 2e-16 ***
## log(P_Dom):feat 0.14505 0.10327 1.405 0.16
## feat:log(P_MM) 0.66526 0.11312 5.881 4.21e-09 ***
## feat:log(P_Trop) -1.73819 0.09673 -17.969 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.3516469)
##
## Null deviance: 6683.9 on 9335 degrees of freedom
## Residual deviance: 3280.2 on 9328 degrees of freedom
## AIC: 16747
##
## Number of Fisher Scoring iterations: 2MM_cross <- subset(oj_cross, brand=="minute.maid")
regcrossm = glm(logmove ~ log(P_Dom)*feat+log(P_MM)*feat+log(P_Trop)*feat, data=MM_cross)
summary(regcrossm)##
## Call:
## glm(formula = logmove ~ log(P_Dom) * feat + log(P_MM) * feat +
## log(P_Trop) * feat, data = MM_cross)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.68138 -0.36949 -0.01947 0.34312 2.72247
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.12742 0.04149 244.096 <2e-16 ***
## log(P_Dom) 0.56164 0.03277 17.141 <2e-16 ***
## feat 1.12989 0.09446 11.962 <2e-16 ***
## log(P_MM) -2.34574 0.04625 -50.722 <2e-16 ***
## log(P_Trop) 0.32894 0.03732 8.813 <2e-16 ***
## log(P_Dom):feat 0.11417 0.06375 1.791 0.0733 .
## feat:log(P_MM) -1.36205 0.08536 -15.956 <2e-16 ***
## feat:log(P_Trop) 0.76155 0.07742 9.837 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.3341122)
##
## Null deviance: 9046.9 on 9335 degrees of freedom
## Residual deviance: 3116.6 on 9328 degrees of freedom
## AIC: 16270
##
## Number of Fisher Scoring iterations: 2dom_cross <- subset(oj_cross, brand=="dominicks")
regcrossd = glm(logmove ~ log(P_Dom)*feat+log(P_MM)*feat+log(P_Trop)*feat, data=dom_cross)
summary(regcrossd)##
## Call:
## glm(formula = logmove ~ log(P_Dom) * feat + log(P_MM) * feat +
## log(P_Trop) * feat, data = dom_cross)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -4.8345 -0.5141 -0.0078 0.4968 3.0611
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.10162 0.06169 163.754 < 2e-16 ***
## log(P_Dom) -2.86900 0.04790 -59.893 < 2e-16 ***
## feat -0.56275 0.11615 -4.845 1.29e-06 ***
## log(P_MM) 0.80506 0.05481 14.689 < 2e-16 ***
## log(P_Trop) -0.26275 0.05159 -5.093 3.59e-07 ***
## log(P_Dom):feat -0.56809 0.08929 -6.362 2.08e-10 ***
## feat:log(P_MM) 1.21816 0.12356 9.859 < 2e-16 ***
## feat:log(P_Trop) 0.71789 0.09660 7.432 1.16e-13 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.6610932)
##
## Null deviance: 13359.9 on 9335 degrees of freedom
## Residual deviance: 6166.7 on 9328 degrees of freedom
## AIC: 22641
##
## Number of Fisher Scoring iterations: 2rownames = c("Q Trop", "Q MM", "Q Dom")
colnames = c("P Trop", "P MM", "P Dom")
Elast_matrix <- matrix(,3,3, dimnames = list(rownames, colnames))
#This code is a hack; it would be much better to do this assignment as a loop; that will show up in the next homework or two
Elast_matrix[1,1] <- coef(regcrosst)["log(P_Trop)"]
Elast_matrix[1,2] <- coef(regcrosst)["log(P_MM)"]
Elast_matrix[1,3] <- coef(regcrosst)["log(P_Dom)"]
Elast_matrix[2,1] <- coef(regcrossm)["log(P_Trop)"]
Elast_matrix[2,2] <- coef(regcrossm)["log(P_MM)"]
Elast_matrix[2,3] <- coef(regcrossm)["log(P_Dom)"]
Elast_matrix[3,1] <- coef(regcrossd)["log(P_Trop)"]
Elast_matrix[3,2] <- coef(regcrossd)["log(P_MM)"]
Elast_matrix[3,3] <- coef(regcrossd)["log(P_Dom)"]
Elast_matrix## P Trop P MM P Dom
## Q Trop -2.1668512 0.2782743 0.1502673
## Q MM 0.3289361 -2.3457404 0.5616439
## Q Dom -0.2627461 0.8050605 -2.8689953# We see that orange juice is a substitute for other orange juice. Minute Maid has twice the cross price elasticity as Dominicks, which makes sense.
###################
## LASSO implementation
###################
# Useful links to tutorials
# https://web.stanford.edu/~hastie/Papers/Glmnet_Vignette.pdf
# http://www4.stat.ncsu.edu/~post/josh/LASSO_Ridge_Elastic_Net_-_Examples.html
#
trop_cross$price <- log(trop_cross$price)
trop_cross$P_MM <- log(trop_cross$P_MM)
trop_cross$P_Dom <- log(trop_cross$P_Dom)
trop_cross$P_Trop <- log(trop_cross$P_Trop)
# NOTE: Must change to reading in the log price directly
x <- as.matrix(trop_cross[ ,5:20])
#x2 <- as.matrix(df[ ,c(2:235, 237)])
y <- as.numeric(as.matrix(trop_cross[ ,4]))
#ydh <- as.double(as.matrix(df[ ,237]))
set.seed(720)
#lasso_v1 <- cv.glmnet(x, y, alpha=1)
lasso_v1 <- glmnet(x, y, alpha=1)
#Results
plot(lasso_v1)
coef(lasso_v1, s=lasso_v1$lambda.min)## 17 x 71 sparse Matrix of class "dgCMatrix"## [[ suppressing 71 column names 's0', 's1', 's2' ... ]]##
## (Intercept) 9.113095 9.3636565 9.5919585 9.7999789 9.9895192 10.162221
## feat . . . . . .
## price . -0.2424893 -0.4634365 -0.6647553 -0.8481895 -1.015328
## AGE60 . . . . . .
## EDUC . . . . . .
## ETHNIC . . . . . .
## INCOME . . . . . .
## HHLARGE . . . . . .
## WORKWOM . . . . . .
## HVAL150 . . . . . .
## SSTRDIST . . . . . .
## SSTRVOL . . . . . .
## CPDIST5 . . . . . .
## CPWVOL5 . . . . . .
## P_Dom . . . . . .
## P_MM . . . . . .
## P_Trop . . . . . .
##
## (Intercept) 10.319581 10.4438187 10.52920524 10.598464158 10.6455464 10.6884154
## feat . 0.0209238 0.07039071 0.114879653 0.1541166 0.1899014
## price -1.167618 -1.2912176 -1.38180665 -1.464394936 -1.5498808 -1.6253536
## AGE60 . . . . . .
## EDUC . . . . . .
## ETHNIC . . . . . .
## INCOME . . . . . .
## HHLARGE . . . . . .
## WORKWOM . . . . . .
## HVAL150 . . . 0.033189385 0.1263768 0.2112843
## SSTRDIST . . . . . .
## SSTRVOL . . . . . .
## CPDIST5 . . . . . .
## CPWVOL5 . . . . . .
## P_Dom . . . . . .
## P_MM . . . . . .
## P_Trop . . . -0.002627611 . .
##
## (Intercept) 10.7274761 10.7703971457 10.8512702 10.862234685 10.8634444942
## feat 0.2225071 0.2521991085 0.2790193 0.302733492 0.3241982639
## price -1.6941215 -1.7568665063 -1.8155678 -1.868905229 -1.9252424074
## AGE60 . . . 0.211324555 0.4322210396
## EDUC . . . . .
## ETHNIC . . . . .
## INCOME . . . . .
## HHLARGE . -0.0521163663 -0.3955450 -0.501861755 -0.5680003552
## WORKWOM . . . . .
## HVAL150 0.2886489 0.3559837608 0.3994133 0.457631866 0.5133662568
## SSTRDIST . . . . .
## SSTRVOL . . . . .
## CPDIST5 . . . . .
## CPWVOL5 . . . . .
## P_Dom . . . . .
## P_MM . . . . .
## P_Trop . -0.0001221817 . -0.003945815 -0.0003646148
##
## (Intercept) 10.8973415361 10.9406370745 10.9791950998 11.0143109284
## feat 0.3436817426 0.3611243888 0.3770009743 0.3914664373
## price -1.9734939612 -2.0180946512 -2.0593292606 -2.0969189669
## AGE60 0.6118646442 0.7659346798 0.9075682969 1.0366413682
## EDUC . . . .
## ETHNIC . . . .
## INCOME . . . .
## HHLARGE -0.6841265782 -0.8185312209 -0.9367821813 -1.0444478405
## WORKWOM . . . .
## HVAL150 0.5669965833 0.6216868723 0.6719642704 0.7177849925
## SSTRDIST . . . .
## SSTRVOL -0.0053368361 -0.0075416240 -0.0094620485 -0.0112096999
## CPDIST5 . . . .
## CPWVOL5 -0.0388784844 -0.0999882498 -0.1557268431 -0.2065157405
## P_Dom . . . .
## P_MM . . . .
## P_Trop -0.0002528143 -0.0008009768 -0.0007472381 -0.0006812219
##
## (Intercept) 11.0463068322 11.0754603015 11.1020238568 11.1262275790
## feat 0.4046468145 0.4166562830 0.4275988635 0.4375693352
## price -2.1311696906 -2.1623776821 -2.1908132418 -2.2167226647
## AGE60 1.1542483722 1.2614075027 1.3590469115 1.4480123043
## EDUC . . . .
## ETHNIC . . . .
## INCOME . . . .
## HHLARGE -1.1425471973 -1.2319316463 -1.3133754266 -1.3875839660
## WORKWOM . . . .
## HVAL150 0.7595353227 0.7975766724 0.8322385336 0.8638211306
## SSTRDIST . . . .
## SSTRVOL -0.0128020518 -0.0142529426 -0.0155749403 -0.0167794954
## CPDIST5 . . . .
## CPWVOL5 -0.2527927315 -0.2949586030 -0.3333785747 -0.3683854220
## P_Dom . . . .
## P_MM . . . .
## P_Trop -0.0006207113 -0.0005655692 -0.0005153256 -0.0004695455
##
## (Intercept) 11.1486714683 11.160758506 11.165655309 11.176423863 11.2486224286
## feat 0.4466778546 0.454531781 0.461423830 0.467709868 0.4734829122
## price -2.2403248009 -2.261990760 -2.285098674 -2.305487267 -2.3237907457
## AGE60 1.5286161326 1.615033438 1.704338548 1.780191473 1.7739203488
## EDUC . . . . .
## ETHNIC . 0.027321572 0.075077360 0.118079216 0.1529456727
## INCOME . . . . .
## HHLARGE -1.4570823575 -1.515792556 -1.568209246 -1.622083356 -1.7589420880
## WORKWOM . . . -0.013546381 -0.1361551644
## HVAL150 0.8923374274 0.924654118 0.957961690 0.988716788 1.0170123570
## SSTRDIST . . . . .
## SSTRVOL -0.0179388731 -0.021746326 -0.028025516 -0.033847942 -0.0404600705
## CPDIST5 . . . . .
## CPWVOL5 -0.4002164697 -0.419383209 -0.427839318 -0.434859834 -0.4362636377
## P_Dom . . . . .
## P_MM . . . . .
## P_Trop -0.0003967452 -0.001891497 -0.001096458 -0.001056376 -0.0009323071
##
## (Intercept) 11.3148574974 11.373497007 11.4150686289 11.44897422 1.165731e+01
## feat 0.4787682213 0.483327510 0.4867086852 0.48928712 4.916038e-01
## price -2.3404204707 -2.355644774 -2.3754685088 -2.39361099 -2.410036e+00
## AGE60 1.7675367884 1.760480539 1.7728719117 1.78649037 1.801561e+00
## EDUC . 0.002981127 0.0748659767 0.13811059 2.192971e-01
## ETHNIC 0.1843194965 0.213038521 0.2379624604 0.26049955 2.676830e-01
## INCOME . . . . -1.751407e-02
## HHLARGE -1.8870904075 -2.003412042 -2.1061171023 -2.19362384 -2.223849e+00
## WORKWOM -0.2470504829 -0.351004220 -0.4556988489 -0.54723985 -6.316530e-01
## HVAL150 1.0423489056 1.064135879 1.0554242249 1.04833764 1.042858e+00
## SSTRDIST . . . . .
## SSTRVOL -0.0464789492 -0.052171640 -0.0588458519 -0.06487980 -7.155204e-02
## CPDIST5 . . . . .
## CPWVOL5 -0.4377308566 -0.438632884 -0.4383352481 -0.43806239 -4.351136e-01
## P_Dom . . 0.0019998538 0.01187433 2.069156e-02
## P_MM . 0.005269569 0.0225912128 0.03714275 4.993092e-02
## P_Trop -0.0008135327 -0.001965341 -0.0002098574 . -9.207667e-05
##
## (Intercept) 1.199013e+01 1.228517e+01 12.6202096559 12.9967141694
## feat 4.938124e-01 4.959368e-01 0.4980080642 0.4999311078
## price -2.424691e+00 -2.437406e+00 -2.4481870288 -2.4578514304
## AGE60 1.853239e+00 1.919592e+00 1.9803534889 2.0425141356
## EDUC 3.311907e-01 4.269284e-01 0.5285200121 0.6275256799
## ETHNIC 2.688311e-01 2.791452e-01 0.2904205046 0.2973114041
## INCOME -4.948575e-02 -7.890468e-02 -0.1137908121 -0.1537724245
## HHLARGE -2.175077e+00 -2.110002e+00 -2.0483096488 -1.9626339224
## WORKWOM -6.812585e-01 -7.060447e-01 -0.7230822137 -0.7300521630
## HVAL150 1.035151e+00 1.033847e+00 1.0365664623 1.0441736485
## SSTRDIST -9.758202e-04 -2.792227e-03 -0.0045537782 -0.0061786858
## SSTRVOL -7.665066e-02 -7.953011e-02 -0.0820556840 -0.0843084609
## CPDIST5 . . 0.0054395009 0.0128226043
## CPWVOL5 -4.376095e-01 -4.456494e-01 -0.4533300325 -0.4601496206
## P_Dom 2.905829e-02 3.703585e-02 0.0445745836 0.0515095088
## P_MM 6.212493e-02 7.395574e-02 0.0855454361 0.0962340765
## P_Trop -8.974930e-06 -3.144655e-05 -0.0002591993 -0.0005213386
##
## (Intercept) 13.321678401 13.6169373957 13.8856011617 14.1331401527
## feat 0.501685143 0.5032834030 0.5047395697 0.5060668619
## price -2.466980443 -2.4752093795 -2.4827167038 -2.4895941148
## AGE60 2.097322348 2.1468616544 2.1919180028 2.2341084029
## EDUC 0.716870319 0.7976826727 0.8711565455 0.9400555036
## ETHNIC 0.305344057 0.3127612488 0.3195630781 0.3254208861
## INCOME -0.188262345 -0.2195852767 -0.2480854815 -0.2743793267
## HHLARGE -1.894323084 -1.8326672547 -1.7766685346 -1.7239771882
## WORKWOM -0.738000445 -0.7454240550 -0.7522132532 -0.7579757350
## HVAL150 1.050026925 1.0555314583 1.0605866662 1.0646060551
## SSTRDIST -0.007672848 -0.0090337864 -0.0102740200 -0.0114046140
## SSTRVOL -0.086335586 -0.0881771783 -0.0898526141 -0.0914021799
## CPDIST5 0.019459596 0.0255046605 0.0310117103 0.0360353442
## CPWVOL5 -0.466528268 -0.4723350686 -0.4776280024 -0.4824571838
## P_Dom 0.057846222 0.0636195333 0.0688799156 0.0736748605
## P_MM 0.106048169 0.1149894751 0.1231367472 0.1305637901
## P_Trop -0.000396378 -0.0003721749 -0.0003408628 -0.0002716806
##
## (Intercept) 14.3558768086 14.558312128 14.7422364587 14.9134867448
## feat 0.5072756906 0.508377028 0.5093804497 0.5102954534
## price -2.4957947788 -2.501469907 -2.5066460405 -2.5113922015
## AGE60 2.2715920914 2.305576859 2.3364692082 2.3656657437
## EDUC 1.0011849867 1.056554498 1.1068395233 1.1545403224
## ETHNIC 0.3310984129 0.336333938 0.3411627297 0.3451275272
## INCOME -0.2980124263 -0.319487611 -0.3389993805 -0.3571912397
## HHLARGE -1.6775812240 -1.635576435 -1.5975345946 -1.5609053639
## WORKWOM -0.7635885235 -0.768748430 -0.7734606334 -0.7773900311
## HVAL150 1.0687297507 1.072585912 1.0761322199 1.0788612574
## SSTRDIST -0.0124349256 -0.013373827 -0.0142297795 -0.0150086639
## SSTRVOL -0.0927937893 -0.094056441 -0.0952027680 -0.0962742031
## CPDIST5 0.0406056368 0.044768878 0.0485606114 0.0520252380
## CPWVOL5 -0.4868574125 -0.490869499 -0.4945320492 -0.4978569263
## P_Dom 0.0780426887 0.082022200 0.0856482774 0.0889531607
## P_MM 0.1373301595 0.143494964 0.1491129649 0.1542311111
## P_Trop -0.0002767043 -0.000256639 -0.0002331071 -0.0001796895
##
## (Intercept) 15.0609994437 15.2031842042 15.3284239716 15.4470835588
## feat 0.5111280114 0.5118877030 0.5125799799 0.5132106604
## price -2.5156413614 -2.5195917434 -2.5231527872 -2.5264171640
## AGE60 2.3904233101 2.4143583095 2.4357099973 2.4558722379
## EDUC 1.1945505214 1.2336650488 1.2682942445 1.3012859026
## ETHNIC 0.3497104754 0.3530136634 0.3564762450 0.3591444113
## INCOME -0.3728328614 -0.3879234181 -0.4012229133 -0.4138254918
## HHLARGE -1.5317805991 -1.5015250539 -1.4758432026 -1.4503450420
## WORKWOM -0.7815454589 -0.7849298117 -0.7880773182 -0.7907931210
## HVAL150 1.0821541912 1.0846115983 1.0869706401 1.0888459766
## SSTRDIST -0.0157228649 -0.0163684045 -0.0169605290 -0.0174966893
## SSTRVOL -0.0971896720 -0.0980759543 -0.0988533321 -0.0995964176
## CPDIST5 0.0551562902 0.0580341408 0.0606405776 0.0630309173
## CPWVOL5 -0.5009501650 -0.5036981504 -0.5062648313 -0.5085455802
## P_Dom 0.0919644135 0.0947071099 0.0972078640 0.0994854797
## P_MM 0.1589031528 0.1631485530 0.1670263819 0.1705525053
## P_Trop -0.0002071088 -0.0001546598 -0.0001420728 -0.0001127566
##
## (Intercept) 15.5503143078 1.564898e+01 1.573379e+01 1.581577e+01
## feat 0.5137863924 5.143098e-01 5.147896e-01 5.152238e-01
## price -2.5293659899 -2.532071e+00 -2.534519e+00 -2.536762e+00
## AGE60 2.4736321609 2.490368e+00 2.505048e+00 2.518905e+00
## EDUC 1.3300547428 1.357463e+00 1.381200e+00 1.403903e+00
## ETHNIC 0.3620688483 3.642723e-01 3.667546e-01 3.685876e-01
## INCOME -0.4247943525 -4.352724e-01 -4.442866e-01 -4.529901e-01
## HHLARGE -1.4292973230 -1.408077e+00 -1.390986e+00 -1.373367e+00
## WORKWOM -0.7933511443 -7.956222e-01 -7.977132e-01 -7.996270e-01
## HVAL150 1.0907448800 1.092304e+00 1.093872e+00 1.095192e+00
## SSTRDIST -0.0179891357 -1.843405e-02 -1.884333e-02 -1.921256e-02
## SSTRVOL -0.1002347542 -1.008538e-01 -1.013736e-01 -1.018890e-01
## CPDIST5 0.0651905204 6.717575e-02 6.896339e-02 7.061231e-02
## CPWVOL5 -0.5107017298 -5.125893e-01 -5.144077e-01 -5.159680e-01
## P_Dom 0.1015625236 1.034533e-01 1.051784e-01 1.067480e-01
## P_MM 0.1737745799 1.767016e-01 1.793781e-01 1.818078e-01
## P_Trop -0.0001075201 -8.833642e-05 -8.167192e-05 -6.906215e-05
##
## (Intercept) 15.8915878237 1.595499e+01 1.601698e+01 1.607474e+01
## feat 0.5156196854 5.159854e-01 5.163134e-01 5.166124e-01
## price -2.5388116170 -2.540655e+00 -2.542347e+00 -2.543897e+00
## AGE60 2.5319006243 2.543050e+00 2.553550e+00 2.563413e+00
## EDUC 1.4252341670 1.443243e+00 1.460452e+00 1.476671e+00
## ETHNIC 0.3701227702 3.720190e-01 3.734181e-01 3.745463e-01
## INCOME -0.4610491533 -4.677952e-01 -4.743776e-01 -4.805154e-01
## HHLARGE -1.3566812965 -1.344023e+00 -1.330700e+00 -1.317911e+00
## WORKWOM -0.8012359345 -8.027412e-01 -8.041976e-01 -8.054245e-01
## HVAL150 1.0962146721 1.097317e+00 1.098301e+00 1.099070e+00
## SSTRDIST -0.0195488833 -1.985869e-02 -2.013820e-02 -2.039232e-02
## SSTRVOL -0.1023668736 -1.027522e-01 -1.031441e-01 -1.035095e-01
## CPDIST5 0.0721175767 7.346348e-02 7.471112e-02 7.585172e-02
## CPWVOL5 -0.5173887479 -5.187953e-01 -5.199714e-01 -5.210365e-01
## P_Dom 0.1081785247 1.094846e-01 1.106719e-01 1.117538e-01
## P_MM 0.1840218620 1.860481e-01 1.878867e-01 1.895609e-01
## P_Trop -0.0000514194 -5.007404e-05 -4.591394e-05 -3.369295e-05
##
## (Intercept) 1.612173e+01 1.616841e+01 1.621241e+01 1.624680e+01
## feat 5.168933e-01 5.171406e-01 5.173661e-01 5.175834e-01
## price -2.545286e+00 -2.546565e+00 -2.547738e+00 -2.548780e+00
## AGE60 2.571745e+00 2.579649e+00 2.587101e+00 2.593239e+00
## EDUC 1.490100e+00 1.503045e+00 1.515321e+00 1.525161e+00
## ETHNIC 3.760091e-01 3.771045e-01 3.779426e-01 3.790799e-01
## INCOME -4.855168e-01 -4.904727e-01 -4.951455e-01 -4.988045e-01
## HHLARGE -1.308759e+00 -1.298788e+00 -1.289005e+00 -1.282587e+00
## WORKWOM -8.065421e-01 -8.076654e-01 -8.086195e-01 -8.094736e-01
## HVAL150 1.099892e+00 1.100654e+00 1.101256e+00 1.101886e+00
## SSTRDIST -2.062589e-02 -2.083770e-02 -2.102977e-02 -2.120441e-02
## SSTRVOL -1.037925e-01 -1.040885e-01 -1.043679e-01 -1.045771e-01
## CPDIST5 7.686156e-02 7.780537e-02 7.867027e-02 7.942291e-02
## CPWVOL5 -5.221256e-01 -5.230141e-01 -5.238093e-01 -5.246465e-01
## P_Dom 1.127425e-01 1.136407e-01 1.144588e-01 1.152076e-01
## P_MM 1.910930e-01 1.924847e-01 1.937505e-01 1.949066e-01
## P_Trop -3.258248e-05 -3.169255e-05 -2.295239e-05 -2.485327e-05
##
## (Intercept) 16.2816298426 1.631503e+01 1.634621e+01 16.3688276262
## feat 0.5177699447 5.179394e-01 5.180938e-01 0.5182506694
## price -2.5497471114 -2.550636e+00 -2.551446e+00 -2.5521512305
## AGE60 2.5991363611 2.604728e+00 2.610005e+00 2.6141752937
## EDUC 1.5348037722 1.544028e+00 1.552753e+00 1.5593354532
## ETHNIC 0.3799701021 3.806160e-01 3.811191e-01 0.3819074572
## INCOME -0.5025019027 -5.060467e-01 -5.093592e-01 -0.5117671572
## HHLARGE -1.2752834201 -1.267867e+00 -1.260719e+00 -1.2566130681
## WORKWOM -0.8103509916 -8.111122e-01 -8.117416e-01 -0.8123433814
## HVAL150 1.1024937020 1.102988e+00 1.103361e+00 1.1037438639
## SSTRDIST -0.0213652851 -2.151073e-02 -2.164298e-02 -0.0217597718
## SSTRVOL -0.1047973837 -1.050102e-01 -1.052089e-01 -0.1053619950
## CPDIST5 0.0801359886 8.079229e-02 8.139255e-02 0.0818968018
## CPWVOL5 -0.5253241112 -5.259170e-01 -5.264537e-01 -0.5270167756
## P_Dom 0.1158871038 1.165056e-01 1.170692e-01 0.1175837818
## P_MM 0.1959607971 1.969183e-01 1.977901e-01 0.1985833188
## P_Trop -0.0000247842 -1.846476e-05 -1.201702e-05 -0.0000186862
##
## (Intercept) 1.639844e+01 1.641180e+01 1.643490e+01 1.644905e+01
## feat 5.183638e-01 5.185106e-01 5.186044e-01 5.187169e-01
## price -2.552846e+00 -2.553388e+00 -2.553980e+00 -2.554451e+00
## AGE60 2.618941e+00 2.621724e+00 2.625286e+00 2.627899e+00
## EDUC 1.567467e+00 1.571209e+00 1.577604e+00 1.581407e+00
## ETHNIC 3.821948e-01 3.830722e-01 3.835637e-01 3.841904e-01
## INCOME -5.149091e-01 -5.163267e-01 -5.187766e-01 -5.202743e-01
## HHLARGE -1.249228e+00 -1.247940e+00 -1.242796e+00 -1.240542e+00
## WORKWOM -8.128909e-01 -8.135300e-01 -8.140436e-01 -8.146006e-01
## HVAL150 1.104057e+00 1.104511e+00 1.104884e+00 1.105280e+00
## SSTRDIST -2.187456e-02 -2.195903e-02 -2.206326e-02 -2.213763e-02
## SSTRVOL -1.055415e-01 -1.056759e-01 -1.057813e-01 -1.059035e-01
## CPDIST5 8.242900e-02 8.279338e-02 8.324512e-02 8.357585e-02
## CPWVOL5 -5.274176e-01 -5.278571e-01 -5.282899e-01 -5.286192e-01
## P_Dom 1.180516e-01 1.184842e-01 1.188679e-01 1.192270e-01
## P_MM 1.993123e-01 1.999527e-01 2.005743e-01 2.011089e-01
## P_Trop -9.897414e-06 -4.332179e-05 -1.586943e-05 -3.519192e-05# Now ready for cross validation version of the object
cvfit <- cv.glmnet(x, y, alpha=1)
#Results
plot(cvfit)
cvfit$lambda.min## [1] 0.0008242127log(cvfit$lambda.min)## [1] -7.101082coef(cvfit, s = "lambda.min")## 17 x 1 sparse Matrix of class "dgCMatrix"
## 1
## (Intercept) 1.644905e+01
## feat 5.187169e-01
## price -2.554451e+00
## AGE60 2.627899e+00
## EDUC 1.581407e+00
## ETHNIC 3.841904e-01
## INCOME -5.202743e-01
## HHLARGE -1.240542e+00
## WORKWOM -8.146006e-01
## HVAL150 1.105280e+00
## SSTRDIST -2.213763e-02
## SSTRVOL -1.059035e-01
## CPDIST5 8.357585e-02
## CPWVOL5 -5.286192e-01
## P_Dom 1.192270e-01
## P_MM 2.011089e-01
## P_Trop -3.519192e-05# Check relative to OLS
reg_lasso <- glm(logmove ~ feat + price + AGE60 + EDUC + ETHNIC + INCOME+ HHLARGE + WORKWOM + HVAL150 + SSTRDIST + SSTRVOL + CPDIST5 + CPWVOL5 + P_MM + P_Dom + P_Trop, data=trop_cross)
summary(reg_lasso)##
## Call:
## glm(formula = logmove ~ feat + price + AGE60 + EDUC + ETHNIC +
## INCOME + HHLARGE + WORKWOM + HVAL150 + SSTRDIST + SSTRVOL +
## CPDIST5 + CPWVOL5 + P_MM + P_Dom + P_Trop, data = trop_cross)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.59850 -0.30183 -0.01105 0.28559 2.81404
##
## Coefficients: (1 not defined because of singularities)
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 16.672684 0.419348 39.759 < 2e-16 ***
## feat 0.519680 0.014537 35.748 < 2e-16 ***
## price -2.559681 0.027930 -91.645 < 2e-16 ***
## AGE60 2.668487 0.161997 16.472 < 2e-16 ***
## EDUC 1.649849 0.129046 12.785 < 2e-16 ***
## ETHNIC 0.385353 0.047286 8.149 4.13e-16 ***
## INCOME -0.544194 0.042310 -12.862 < 2e-16 ***
## HHLARGE -1.182697 0.293007 -4.036 5.47e-05 ***
## WORKWOM -0.816628 0.186575 -4.377 1.22e-05 ***
## HVAL150 1.104607 0.053199 20.764 < 2e-16 ***
## SSTRDIST -0.023005 0.001868 -12.317 < 2e-16 ***
## SSTRVOL -0.107334 0.012439 -8.629 < 2e-16 ***
## CPDIST5 0.087552 0.007996 10.950 < 2e-16 ***
## CPWVOL5 -0.532020 0.032704 -16.268 < 2e-16 ***
## P_MM 0.206762 0.028859 7.165 8.40e-13 ***
## P_Dom 0.122861 0.022711 5.410 6.47e-08 ***
## P_Trop NA NA NA NA
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.224841)
##
## Null deviance: 6683.9 on 9335 degrees of freedom
## Residual deviance: 2095.5 on 9320 degrees of freedom
## AIC: 12580
##
## Number of Fisher Scoring iterations: 2plot(cvfit, xvar = "lambda", label = TRUE)## Warning in plot.window(...): "xvar" is not a graphical parameter## Warning in plot.window(...): "label" is not a graphical parameter## Warning in plot.xy(xy, type, ...): "xvar" is not a graphical parameter## Warning in plot.xy(xy, type, ...): "label" is not a graphical parameter## Warning in axis(side = side, at = at, labels = labels, ...): "xvar" is not a
## graphical parameter## Warning in axis(side = side, at = at, labels = labels, ...): "label" is not a
## graphical parameter## Warning in axis(side = side, at = at, labels = labels, ...): "xvar" is not a
## graphical parameter## Warning in axis(side = side, at = at, labels = labels, ...): "label" is not a
## graphical parameter## Warning in box(...): "xvar" is not a graphical parameter## Warning in box(...): "label" is not a graphical parameter## Warning in title(...): "xvar" is not a graphical parameter## Warning in title(...): "label" is not a graphical parameter
plot(lasso_v1, xvar = "dev", label = TRUE)
#plot(lasso_v1$glmnet.fit, xvar="lambda", label=TRUE) # There is a term
#lasso_v1$lambda.min
#lasso_v1$lambda.1se
#coef(lasso_v1, s=cv.lasso$lambda.min)
coef(lasso_v1, s=lasso_v1$lambda.min)## 17 x 71 sparse Matrix of class "dgCMatrix"## [[ suppressing 71 column names 's0', 's1', 's2' ... ]]##
## (Intercept) 9.113095 9.3636565 9.5919585 9.7999789 9.9895192 10.162221
## feat . . . . . .
## price . -0.2424893 -0.4634365 -0.6647553 -0.8481895 -1.015328
## AGE60 . . . . . .
## EDUC . . . . . .
## ETHNIC . . . . . .
## INCOME . . . . . .
## HHLARGE . . . . . .
## WORKWOM . . . . . .
## HVAL150 . . . . . .
## SSTRDIST . . . . . .
## SSTRVOL . . . . . .
## CPDIST5 . . . . . .
## CPWVOL5 . . . . . .
## P_Dom . . . . . .
## P_MM . . . . . .
## P_Trop . . . . . .
##
## (Intercept) 10.319581 10.4438187 10.52920524 10.598464158 10.6455464 10.6884154
## feat . 0.0209238 0.07039071 0.114879653 0.1541166 0.1899014
## price -1.167618 -1.2912176 -1.38180665 -1.464394936 -1.5498808 -1.6253536
## AGE60 . . . . . .
## EDUC . . . . . .
## ETHNIC . . . . . .
## INCOME . . . . . .
## HHLARGE . . . . . .
## WORKWOM . . . . . .
## HVAL150 . . . 0.033189385 0.1263768 0.2112843
## SSTRDIST . . . . . .
## SSTRVOL . . . . . .
## CPDIST5 . . . . . .
## CPWVOL5 . . . . . .
## P_Dom . . . . . .
## P_MM . . . . . .
## P_Trop . . . -0.002627611 . .
##
## (Intercept) 10.7274761 10.7703971457 10.8512702 10.862234685 10.8634444942
## feat 0.2225071 0.2521991085 0.2790193 0.302733492 0.3241982639
## price -1.6941215 -1.7568665063 -1.8155678 -1.868905229 -1.9252424074
## AGE60 . . . 0.211324555 0.4322210396
## EDUC . . . . .
## ETHNIC . . . . .
## INCOME . . . . .
## HHLARGE . -0.0521163663 -0.3955450 -0.501861755 -0.5680003552
## WORKWOM . . . . .
## HVAL150 0.2886489 0.3559837608 0.3994133 0.457631866 0.5133662568
## SSTRDIST . . . . .
## SSTRVOL . . . . .
## CPDIST5 . . . . .
## CPWVOL5 . . . . .
## P_Dom . . . . .
## P_MM . . . . .
## P_Trop . -0.0001221817 . -0.003945815 -0.0003646148
##
## (Intercept) 10.8973415361 10.9406370745 10.9791950998 11.0143109284
## feat 0.3436817426 0.3611243888 0.3770009743 0.3914664373
## price -1.9734939612 -2.0180946512 -2.0593292606 -2.0969189669
## AGE60 0.6118646442 0.7659346798 0.9075682969 1.0366413682
## EDUC . . . .
## ETHNIC . . . .
## INCOME . . . .
## HHLARGE -0.6841265782 -0.8185312209 -0.9367821813 -1.0444478405
## WORKWOM . . . .
## HVAL150 0.5669965833 0.6216868723 0.6719642704 0.7177849925
## SSTRDIST . . . .
## SSTRVOL -0.0053368361 -0.0075416240 -0.0094620485 -0.0112096999
## CPDIST5 . . . .
## CPWVOL5 -0.0388784844 -0.0999882498 -0.1557268431 -0.2065157405
## P_Dom . . . .
## P_MM . . . .
## P_Trop -0.0002528143 -0.0008009768 -0.0007472381 -0.0006812219
##
## (Intercept) 11.0463068322 11.0754603015 11.1020238568 11.1262275790
## feat 0.4046468145 0.4166562830 0.4275988635 0.4375693352
## price -2.1311696906 -2.1623776821 -2.1908132418 -2.2167226647
## AGE60 1.1542483722 1.2614075027 1.3590469115 1.4480123043
## EDUC . . . .
## ETHNIC . . . .
## INCOME . . . .
## HHLARGE -1.1425471973 -1.2319316463 -1.3133754266 -1.3875839660
## WORKWOM . . . .
## HVAL150 0.7595353227 0.7975766724 0.8322385336 0.8638211306
## SSTRDIST . . . .
## SSTRVOL -0.0128020518 -0.0142529426 -0.0155749403 -0.0167794954
## CPDIST5 . . . .
## CPWVOL5 -0.2527927315 -0.2949586030 -0.3333785747 -0.3683854220
## P_Dom . . . .
## P_MM . . . .
## P_Trop -0.0006207113 -0.0005655692 -0.0005153256 -0.0004695455
##
## (Intercept) 11.1486714683 11.160758506 11.165655309 11.176423863 11.2486224286
## feat 0.4466778546 0.454531781 0.461423830 0.467709868 0.4734829122
## price -2.2403248009 -2.261990760 -2.285098674 -2.305487267 -2.3237907457
## AGE60 1.5286161326 1.615033438 1.704338548 1.780191473 1.7739203488
## EDUC . . . . .
## ETHNIC . 0.027321572 0.075077360 0.118079216 0.1529456727
## INCOME . . . . .
## HHLARGE -1.4570823575 -1.515792556 -1.568209246 -1.622083356 -1.7589420880
## WORKWOM . . . -0.013546381 -0.1361551644
## HVAL150 0.8923374274 0.924654118 0.957961690 0.988716788 1.0170123570
## SSTRDIST . . . . .
## SSTRVOL -0.0179388731 -0.021746326 -0.028025516 -0.033847942 -0.0404600705
## CPDIST5 . . . . .
## CPWVOL5 -0.4002164697 -0.419383209 -0.427839318 -0.434859834 -0.4362636377
## P_Dom . . . . .
## P_MM . . . . .
## P_Trop -0.0003967452 -0.001891497 -0.001096458 -0.001056376 -0.0009323071
##
## (Intercept) 11.3148574974 11.373497007 11.4150686289 11.44897422 1.165731e+01
## feat 0.4787682213 0.483327510 0.4867086852 0.48928712 4.916038e-01
## price -2.3404204707 -2.355644774 -2.3754685088 -2.39361099 -2.410036e+00
## AGE60 1.7675367884 1.760480539 1.7728719117 1.78649037 1.801561e+00
## EDUC . 0.002981127 0.0748659767 0.13811059 2.192971e-01
## ETHNIC 0.1843194965 0.213038521 0.2379624604 0.26049955 2.676830e-01
## INCOME . . . . -1.751407e-02
## HHLARGE -1.8870904075 -2.003412042 -2.1061171023 -2.19362384 -2.223849e+00
## WORKWOM -0.2470504829 -0.351004220 -0.4556988489 -0.54723985 -6.316530e-01
## HVAL150 1.0423489056 1.064135879 1.0554242249 1.04833764 1.042858e+00
## SSTRDIST . . . . .
## SSTRVOL -0.0464789492 -0.052171640 -0.0588458519 -0.06487980 -7.155204e-02
## CPDIST5 . . . . .
## CPWVOL5 -0.4377308566 -0.438632884 -0.4383352481 -0.43806239 -4.351136e-01
## P_Dom . . 0.0019998538 0.01187433 2.069156e-02
## P_MM . 0.005269569 0.0225912128 0.03714275 4.993092e-02
## P_Trop -0.0008135327 -0.001965341 -0.0002098574 . -9.207667e-05
##
## (Intercept) 1.199013e+01 1.228517e+01 12.6202096559 12.9967141694
## feat 4.938124e-01 4.959368e-01 0.4980080642 0.4999311078
## price -2.424691e+00 -2.437406e+00 -2.4481870288 -2.4578514304
## AGE60 1.853239e+00 1.919592e+00 1.9803534889 2.0425141356
## EDUC 3.311907e-01 4.269284e-01 0.5285200121 0.6275256799
## ETHNIC 2.688311e-01 2.791452e-01 0.2904205046 0.2973114041
## INCOME -4.948575e-02 -7.890468e-02 -0.1137908121 -0.1537724245
## HHLARGE -2.175077e+00 -2.110002e+00 -2.0483096488 -1.9626339224
## WORKWOM -6.812585e-01 -7.060447e-01 -0.7230822137 -0.7300521630
## HVAL150 1.035151e+00 1.033847e+00 1.0365664623 1.0441736485
## SSTRDIST -9.758202e-04 -2.792227e-03 -0.0045537782 -0.0061786858
## SSTRVOL -7.665066e-02 -7.953011e-02 -0.0820556840 -0.0843084609
## CPDIST5 . . 0.0054395009 0.0128226043
## CPWVOL5 -4.376095e-01 -4.456494e-01 -0.4533300325 -0.4601496206
## P_Dom 2.905829e-02 3.703585e-02 0.0445745836 0.0515095088
## P_MM 6.212493e-02 7.395574e-02 0.0855454361 0.0962340765
## P_Trop -8.974930e-06 -3.144655e-05 -0.0002591993 -0.0005213386
##
## (Intercept) 13.321678401 13.6169373957 13.8856011617 14.1331401527
## feat 0.501685143 0.5032834030 0.5047395697 0.5060668619
## price -2.466980443 -2.4752093795 -2.4827167038 -2.4895941148
## AGE60 2.097322348 2.1468616544 2.1919180028 2.2341084029
## EDUC 0.716870319 0.7976826727 0.8711565455 0.9400555036
## ETHNIC 0.305344057 0.3127612488 0.3195630781 0.3254208861
## INCOME -0.188262345 -0.2195852767 -0.2480854815 -0.2743793267
## HHLARGE -1.894323084 -1.8326672547 -1.7766685346 -1.7239771882
## WORKWOM -0.738000445 -0.7454240550 -0.7522132532 -0.7579757350
## HVAL150 1.050026925 1.0555314583 1.0605866662 1.0646060551
## SSTRDIST -0.007672848 -0.0090337864 -0.0102740200 -0.0114046140
## SSTRVOL -0.086335586 -0.0881771783 -0.0898526141 -0.0914021799
## CPDIST5 0.019459596 0.0255046605 0.0310117103 0.0360353442
## CPWVOL5 -0.466528268 -0.4723350686 -0.4776280024 -0.4824571838
## P_Dom 0.057846222 0.0636195333 0.0688799156 0.0736748605
## P_MM 0.106048169 0.1149894751 0.1231367472 0.1305637901
## P_Trop -0.000396378 -0.0003721749 -0.0003408628 -0.0002716806
##
## (Intercept) 14.3558768086 14.558312128 14.7422364587 14.9134867448
## feat 0.5072756906 0.508377028 0.5093804497 0.5102954534
## price -2.4957947788 -2.501469907 -2.5066460405 -2.5113922015
## AGE60 2.2715920914 2.305576859 2.3364692082 2.3656657437
## EDUC 1.0011849867 1.056554498 1.1068395233 1.1545403224
## ETHNIC 0.3310984129 0.336333938 0.3411627297 0.3451275272
## INCOME -0.2980124263 -0.319487611 -0.3389993805 -0.3571912397
## HHLARGE -1.6775812240 -1.635576435 -1.5975345946 -1.5609053639
## WORKWOM -0.7635885235 -0.768748430 -0.7734606334 -0.7773900311
## HVAL150 1.0687297507 1.072585912 1.0761322199 1.0788612574
## SSTRDIST -0.0124349256 -0.013373827 -0.0142297795 -0.0150086639
## SSTRVOL -0.0927937893 -0.094056441 -0.0952027680 -0.0962742031
## CPDIST5 0.0406056368 0.044768878 0.0485606114 0.0520252380
## CPWVOL5 -0.4868574125 -0.490869499 -0.4945320492 -0.4978569263
## P_Dom 0.0780426887 0.082022200 0.0856482774 0.0889531607
## P_MM 0.1373301595 0.143494964 0.1491129649 0.1542311111
## P_Trop -0.0002767043 -0.000256639 -0.0002331071 -0.0001796895
##
## (Intercept) 15.0609994437 15.2031842042 15.3284239716 15.4470835588
## feat 0.5111280114 0.5118877030 0.5125799799 0.5132106604
## price -2.5156413614 -2.5195917434 -2.5231527872 -2.5264171640
## AGE60 2.3904233101 2.4143583095 2.4357099973 2.4558722379
## EDUC 1.1945505214 1.2336650488 1.2682942445 1.3012859026
## ETHNIC 0.3497104754 0.3530136634 0.3564762450 0.3591444113
## INCOME -0.3728328614 -0.3879234181 -0.4012229133 -0.4138254918
## HHLARGE -1.5317805991 -1.5015250539 -1.4758432026 -1.4503450420
## WORKWOM -0.7815454589 -0.7849298117 -0.7880773182 -0.7907931210
## HVAL150 1.0821541912 1.0846115983 1.0869706401 1.0888459766
## SSTRDIST -0.0157228649 -0.0163684045 -0.0169605290 -0.0174966893
## SSTRVOL -0.0971896720 -0.0980759543 -0.0988533321 -0.0995964176
## CPDIST5 0.0551562902 0.0580341408 0.0606405776 0.0630309173
## CPWVOL5 -0.5009501650 -0.5036981504 -0.5062648313 -0.5085455802
## P_Dom 0.0919644135 0.0947071099 0.0972078640 0.0994854797
## P_MM 0.1589031528 0.1631485530 0.1670263819 0.1705525053
## P_Trop -0.0002071088 -0.0001546598 -0.0001420728 -0.0001127566
##
## (Intercept) 15.5503143078 1.564898e+01 1.573379e+01 1.581577e+01
## feat 0.5137863924 5.143098e-01 5.147896e-01 5.152238e-01
## price -2.5293659899 -2.532071e+00 -2.534519e+00 -2.536762e+00
## AGE60 2.4736321609 2.490368e+00 2.505048e+00 2.518905e+00
## EDUC 1.3300547428 1.357463e+00 1.381200e+00 1.403903e+00
## ETHNIC 0.3620688483 3.642723e-01 3.667546e-01 3.685876e-01
## INCOME -0.4247943525 -4.352724e-01 -4.442866e-01 -4.529901e-01
## HHLARGE -1.4292973230 -1.408077e+00 -1.390986e+00 -1.373367e+00
## WORKWOM -0.7933511443 -7.956222e-01 -7.977132e-01 -7.996270e-01
## HVAL150 1.0907448800 1.092304e+00 1.093872e+00 1.095192e+00
## SSTRDIST -0.0179891357 -1.843405e-02 -1.884333e-02 -1.921256e-02
## SSTRVOL -0.1002347542 -1.008538e-01 -1.013736e-01 -1.018890e-01
## CPDIST5 0.0651905204 6.717575e-02 6.896339e-02 7.061231e-02
## CPWVOL5 -0.5107017298 -5.125893e-01 -5.144077e-01 -5.159680e-01
## P_Dom 0.1015625236 1.034533e-01 1.051784e-01 1.067480e-01
## P_MM 0.1737745799 1.767016e-01 1.793781e-01 1.818078e-01
## P_Trop -0.0001075201 -8.833642e-05 -8.167192e-05 -6.906215e-05
##
## (Intercept) 15.8915878237 1.595499e+01 1.601698e+01 1.607474e+01
## feat 0.5156196854 5.159854e-01 5.163134e-01 5.166124e-01
## price -2.5388116170 -2.540655e+00 -2.542347e+00 -2.543897e+00
## AGE60 2.5319006243 2.543050e+00 2.553550e+00 2.563413e+00
## EDUC 1.4252341670 1.443243e+00 1.460452e+00 1.476671e+00
## ETHNIC 0.3701227702 3.720190e-01 3.734181e-01 3.745463e-01
## INCOME -0.4610491533 -4.677952e-01 -4.743776e-01 -4.805154e-01
## HHLARGE -1.3566812965 -1.344023e+00 -1.330700e+00 -1.317911e+00
## WORKWOM -0.8012359345 -8.027412e-01 -8.041976e-01 -8.054245e-01
## HVAL150 1.0962146721 1.097317e+00 1.098301e+00 1.099070e+00
## SSTRDIST -0.0195488833 -1.985869e-02 -2.013820e-02 -2.039232e-02
## SSTRVOL -0.1023668736 -1.027522e-01 -1.031441e-01 -1.035095e-01
## CPDIST5 0.0721175767 7.346348e-02 7.471112e-02 7.585172e-02
## CPWVOL5 -0.5173887479 -5.187953e-01 -5.199714e-01 -5.210365e-01
## P_Dom 0.1081785247 1.094846e-01 1.106719e-01 1.117538e-01
## P_MM 0.1840218620 1.860481e-01 1.878867e-01 1.895609e-01
## P_Trop -0.0000514194 -5.007404e-05 -4.591394e-05 -3.369295e-05
##
## (Intercept) 1.612173e+01 1.616841e+01 1.621241e+01 1.624680e+01
## feat 5.168933e-01 5.171406e-01 5.173661e-01 5.175834e-01
## price -2.545286e+00 -2.546565e+00 -2.547738e+00 -2.548780e+00
## AGE60 2.571745e+00 2.579649e+00 2.587101e+00 2.593239e+00
## EDUC 1.490100e+00 1.503045e+00 1.515321e+00 1.525161e+00
## ETHNIC 3.760091e-01 3.771045e-01 3.779426e-01 3.790799e-01
## INCOME -4.855168e-01 -4.904727e-01 -4.951455e-01 -4.988045e-01
## HHLARGE -1.308759e+00 -1.298788e+00 -1.289005e+00 -1.282587e+00
## WORKWOM -8.065421e-01 -8.076654e-01 -8.086195e-01 -8.094736e-01
## HVAL150 1.099892e+00 1.100654e+00 1.101256e+00 1.101886e+00
## SSTRDIST -2.062589e-02 -2.083770e-02 -2.102977e-02 -2.120441e-02
## SSTRVOL -1.037925e-01 -1.040885e-01 -1.043679e-01 -1.045771e-01
## CPDIST5 7.686156e-02 7.780537e-02 7.867027e-02 7.942291e-02
## CPWVOL5 -5.221256e-01 -5.230141e-01 -5.238093e-01 -5.246465e-01
## P_Dom 1.127425e-01 1.136407e-01 1.144588e-01 1.152076e-01
## P_MM 1.910930e-01 1.924847e-01 1.937505e-01 1.949066e-01
## P_Trop -3.258248e-05 -3.169255e-05 -2.295239e-05 -2.485327e-05
##
## (Intercept) 16.2816298426 1.631503e+01 1.634621e+01 16.3688276262
## feat 0.5177699447 5.179394e-01 5.180938e-01 0.5182506694
## price -2.5497471114 -2.550636e+00 -2.551446e+00 -2.5521512305
## AGE60 2.5991363611 2.604728e+00 2.610005e+00 2.6141752937
## EDUC 1.5348037722 1.544028e+00 1.552753e+00 1.5593354532
## ETHNIC 0.3799701021 3.806160e-01 3.811191e-01 0.3819074572
## INCOME -0.5025019027 -5.060467e-01 -5.093592e-01 -0.5117671572
## HHLARGE -1.2752834201 -1.267867e+00 -1.260719e+00 -1.2566130681
## WORKWOM -0.8103509916 -8.111122e-01 -8.117416e-01 -0.8123433814
## HVAL150 1.1024937020 1.102988e+00 1.103361e+00 1.1037438639
## SSTRDIST -0.0213652851 -2.151073e-02 -2.164298e-02 -0.0217597718
## SSTRVOL -0.1047973837 -1.050102e-01 -1.052089e-01 -0.1053619950
## CPDIST5 0.0801359886 8.079229e-02 8.139255e-02 0.0818968018
## CPWVOL5 -0.5253241112 -5.259170e-01 -5.264537e-01 -0.5270167756
## P_Dom 0.1158871038 1.165056e-01 1.170692e-01 0.1175837818
## P_MM 0.1959607971 1.969183e-01 1.977901e-01 0.1985833188
## P_Trop -0.0000247842 -1.846476e-05 -1.201702e-05 -0.0000186862
##
## (Intercept) 1.639844e+01 1.641180e+01 1.643490e+01 1.644905e+01
## feat 5.183638e-01 5.185106e-01 5.186044e-01 5.187169e-01
## price -2.552846e+00 -2.553388e+00 -2.553980e+00 -2.554451e+00
## AGE60 2.618941e+00 2.621724e+00 2.625286e+00 2.627899e+00
## EDUC 1.567467e+00 1.571209e+00 1.577604e+00 1.581407e+00
## ETHNIC 3.821948e-01 3.830722e-01 3.835637e-01 3.841904e-01
## INCOME -5.149091e-01 -5.163267e-01 -5.187766e-01 -5.202743e-01
## HHLARGE -1.249228e+00 -1.247940e+00 -1.242796e+00 -1.240542e+00
## WORKWOM -8.128909e-01 -8.135300e-01 -8.140436e-01 -8.146006e-01
## HVAL150 1.104057e+00 1.104511e+00 1.104884e+00 1.105280e+00
## SSTRDIST -2.187456e-02 -2.195903e-02 -2.206326e-02 -2.213763e-02
## SSTRVOL -1.055415e-01 -1.056759e-01 -1.057813e-01 -1.059035e-01
## CPDIST5 8.242900e-02 8.279338e-02 8.324512e-02 8.357585e-02
## CPWVOL5 -5.274176e-01 -5.278571e-01 -5.282899e-01 -5.286192e-01
## P_Dom 1.180516e-01 1.184842e-01 1.188679e-01 1.192270e-01
## P_MM 1.993123e-01 1.999527e-01 2.005743e-01 2.011089e-01
## P_Trop -9.897414e-06 -4.332179e-05 -1.586943e-05 -3.519192e-05# The key this here is that the week variable is formatted as a date variable. This provides R with some information that it is a panel dataset
#create a date and sequence accompanying the dates within the dataframe then use lag operaters to make progress on it.Note that the echo = FALSE parameter was added to the code chunk to prevent printing of the R code that generated the plot.