-
Notifications
You must be signed in to change notification settings - Fork 0
/
Case Comp Time Series (1).R
383 lines (262 loc) · 12.5 KB
/
Case Comp Time Series (1).R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
library(fpp2)
library(tidyverse)
library(readxl)
library(forecast)
library(ggplot2)
library(dplyr)
library(reshape2)
library(janitor)
library(tibble)
#read data
getwd()
setwd('C:/Users/kesha/Desktop/MSBA Keshav/Case Competitions/MSBA Case Competition')
#elec = read.table("dtwnLAozone.txt", skip=2, header = TRUE)
elec = read_excel('UPC Utility Data 2018-2021.xlsx', sheet="Electricity", range = "A6:AY232")
class(elec)
elec
elec_ts_clean = data.frame(t(elec[c(-1,-224,-225),-1:-2]))
elec_ts_clean
head(elec_ts_clean)
dim(elec_ts_clean)
#elec_ts_clean[1,223] = "TOT"
elec_ts_TOT = elec_ts_clean[-1,223]
elec_ts_TOT = as.numeric(elec_ts_TOT)
class(elec_ts_TOT)
length(elec_ts_TOT)
# Create a time series object for data
elec_ts_TOT.ts = ts(elec_ts_TOT,start=c(2018,1),frequency=12)
autoplot(elec_ts_TOT.ts)
length(elec_ts_TOT.ts)
# Divide data into training and testing sets and call y.train and y.test
y.train = window(elec_ts_TOT.ts, end=c(2020,3))
y.train
y.test = window(elec_ts_TOT.ts, start=c(2020,4))
y.test
length(y.test)
# Naive models: non-seasonal naive model and seasonal naive model
# Naive model prediction for time t+1 = Actual data at time t
# Build a model on training set and predict on the testing set
?naive
M1 = naive(y.train, h = length(y.test))
M1
# Fitted values: predicted values on the training set
M1$fitted
# Forecast on the testing set
M1$mean
# Plot all data (train + test) and overlay fitted values and forecasts
autoplot(elec_ts_TOT.ts) + autolayer(M1$fitted) + autolayer(M1$mean,size=2)+
theme_bw()
#Accuracy metrics on both training and testing sets:
accuracy(M1,y.test)
# Seasonal pattern was not modeled. Instead of naive we should use snaive.
# snaive = seasonal naive model:
M2 = snaive(y.train, h = length(y.test), level=95)
autoplot(elec_ts_TOT.ts) + autolayer(M2$fitted) + autolayer(M2$mean,size=2)+
theme_bw()
#autoplot(y) + autoplot(
tsdisplay(elec_ts_TOT.ts)
ndiffs(elec_ts_TOT.ts)
nsdiffs(elec_ts_TOT.ts)
?ARIMA
?ma
elec.trend = ma(elec_ts_TOT.ts, order = 12)
autoplot(elec_ts_TOT.ts) + autolayer(elec.trend)
elec_ts_TOT.ts
elec_ts_clean = janitor::row_to_names(elec_ts_clean, row_number=1) %>% clean_names()
elec_ts_clean
length(unique(colnames(elec_ts_clean)))
length(colnames(elec_ts_clean))
colnames(elec_ts_clean)
df = tibble(x=colnames(elec_ts_clean))
dplyr::count(df, x,sort=TRUE)
row.names(elec_ts_clean)
colnames(elec_ts_clean)
#elec_ts_clean[elec_ts_clean$]
#colnames(elec_ts_clean) = elec_ts_clean[1, ]
elec_ts_clean$adm <- as.numeric(elec_ts_clean$adm)
#steps: get the header put in... done with clean names?
elec_ts_clean = elec_ts_clean %>% mutate(Ramp=c(rep(0,27),1:21))
elec_ts_clean = elec_ts_clean %>% mutate(Bump=c(rep(0,27),rep(1,21)))
elec_ts_clean$Month = 1:(dim(elec_ts_clean)[1])
elec_ts_clean = elec_ts_clean %>% mutate(MonthShort=c('Jan','Feb','March','April','May','June','July','Aug','Sep','Oct','Nov','Dec',
'Jan','Feb','March','April','May','June','July','Aug','Sep','Oct','Nov','Dec',
'Jan','Feb','March','April','May','June','July','Aug','Sep','Oct','Nov','Dec',
'Jan','Feb','March','April','May','June','July','Aug','Sep','Oct','Nov','Dec'))
elec_ts_clean$Bump
elec_ts_clean$Ramp
elec_ts_clean$Month
elec_ts_clean$MonthShort
#create lag1
elec_ts_clean$adm_Lag1 <- elec_ts_clean$adm %>% lag(1)
#M11 creation - example to make loop
elec_ts_clean$M11ADM
M11=lm(adm ~ Bump + Ramp + Month + MonthShort + adm_Lag1,data=elec_ts_clean)
elec_ts_clean$M11ADM[2:48] = M11$fitted.values
elec_ts_clean$M11ADM
elec_ts_clean %>% ggplot(mapping=aes(x=Month,y=adm)) + geom_point() + theme_bw()+
geom_line(aes(x=Month,y=adm)) +
geom_line(mapping=aes(x=Month,y=M11ADM),col="red",lwd=1)
#M11 test creation end - commented out section
summary(M11)[9] #R^2 reference
M11$coefficients[3]
plot(M11$residuals)
#answer -50.41905
dim(elec_ts_clean)
colnames(elec_ts_clean)[1:223]
#create data frame to store results, row names are the bldg codes and coefficients are the post-covid trend coefficient (Ramp)
elec_results = data.frame(colnames(elec_ts_clean[1:223]),coefficient_value=0,row.names = colnames(elec_ts_clean[1:223]))
elec_results = elec_results[-1]
#colnames(elec_results)
elec_results
adm_Lag1 <- elec_ts_clean['adm'] %>% lag(1)
for (i in colnames(elec_ts_clean)[1:223]) {
elec_ts_clean[i] <- as.numeric(unlist(elec_ts_clean[i]))
lag_name <- paste("i", "_Lag1", sep = "")
elec_ts_clean[lag_name] = elec_ts_clean[i] %>% lag(1)
M_i = lm(unlist(elec_ts_clean[i]) ~ Ramp + Bump + Month + MonthShort + i_Lag1,data=elec_ts_clean)
elec_results[i,"coefficient_value"] = M_i$coefficients[3]
elec_results[i,"Model R Squared"] = summary(M_i)[9]
elec_results[i,"mean"] = mean(unlist(elec_ts_clean[i]))
}
elec_results
write.csv(elec_results,"electricity_mean_trend.csv")
#mean(unlist(elec_ts_clean["adm"]))
####debug negative R2
elec_ts_clean$dmv <- as.numeric(elec_ts_clean$dmv)
elec_ts_clean$dmv_Lag1 <- elec_ts_clean$dmv %>% lag(1)
M11=lm(dmv ~ Bump + Ramp + Month + MonthShort + dmv_Lag1,data=elec_ts_clean)
elec_ts_clean %>% ggplot(mapping=aes(x=Month,y=dmv)) + geom_point() + theme_bw()+
geom_line(aes(x=Month,y=dmv)) +
geom_line(mapping=aes(x=Month,y=M11ADM),col="red",lwd=1)
summary(M11)
########end debug negative R2
#colnames(elec_ts_clean)
#for (i in colnames(elec_ts_clean)[1:223]){
#print(dim(elec_ts_clean[i])[1])
#print(count(elec_ts_clean[i]))
# print(i)
#}
#length(elec_ts_clean$adm)
#M_adm = lm(adm ~ Bump + Month,data=elec_ts_clean)
#elec_results["adm","coefficient_value"] = M_adm$coefficients[3]
#test to see if i can find address of cell using row and col name
#elec_results["adm","coefficient_value"]
#elec_results
#elec_results <- elec_results %>%
# column_to_rownames('colnames.elec_ts_clean.')
#for (i in colnames(elec_ts_clean)){
#print(dim(elec_ts_clean[i])[1])
# print(count(elec_ts_clean[i]))
#print(i)
#}
#remove outliers
#elec_results_clean = elec_results[elec_results$mean <= 10000000,]
#plot(elec_results_clean$coefficient_value, elec_results_clean$mean, main="Scatterplot",
# xlab="coefficient", ylab="mean")
########
# Electricity section attempt
########
elec = read_excel('UPC Utility Data 2018-2021.xlsx', sheet="Electricity", range = "A6:AY232")
elec_clean = data.frame(t(elec[c(-1,-224,-225),-1:-2]))
elec_clean = janitor::row_to_names(elec_clean, row_number=1) %>% clean_names()
elec_clean
#create data frame to store results, row names are the bldg codes and coefficients are the post-covid trend coefficient (Ramp)
elec_results = data.frame(colnames(elec_clean[1:223]),coefficient_value=0,row.names = colnames(elec_clean[1:223]))
elec_results = elec_results[-1]
#colnames(elec_results)
elec_results
elec_clean = elec_clean %>% mutate(Ramp=c(rep(0,27),1:21))
elec_clean = elec_clean %>% mutate(Bump=c(rep(0,27),rep(1,21)))
elec_clean$Month = 1:(dim(elec_clean)[1])
elec_clean = elec_clean %>% mutate(MonthShort=c('Jan','Feb','March','April','May','June','July','Aug','Sep','Oct','Nov','Dec',
'Jan','Feb','March','April','May','June','July','Aug','Sep','Oct','Nov','Dec',
'Jan','Feb','March','April','May','June','July','Aug','Sep','Oct','Nov','Dec',
'Jan','Feb','March','April','May','June','July','Aug','Sep','Oct','Nov','Dec'))
#elec_clean$adm_Lag1 <- elec_clean$adm %>% lag(1)
for (i in colnames(elec_clean)[1:223]) {
elec_clean[i] <- as.numeric(unlist(elec_clean[i]))
M_i = lm(unlist(elec_clean[i]) ~ Ramp + Bump + Month + MonthShort,data=elec_clean)
elec_results[i,"coefficient_value"] = M_i$coefficients[3]
elec_results[i,"mean"] = mean(unlist(elec_clean[i]))
}
elec_results
write.csv(elec_results,"electricity_mean_trend.csv")
################
# nat_gas section start
################
nat_gas = read_excel('UPC Utility Data 2018-2021.xlsx', sheet="Natural Gas", range = "A3:AY181")
nat_gas_clean = data.frame(t(nat_gas[c(-179,-180),-1:-2]))
head(nat_gas_clean)
dim(nat_gas_clean)
nat_gas_clean[1:175]
#janitor to get rid of auto generated column names and put proper column names on
nat_gas_clean = janitor::row_to_names(nat_gas_clean, row_number=1) %>% clean_names()
dim(nat_gas_clean)
nat_gas_clean = nat_gas_clean[c(-176,-177,-178)]
nat_gas_clean
nat_gas_clean = nat_gas_clean %>% mutate(Ramp=c(rep(0,27),1:21))
nat_gas_clean = nat_gas_clean %>% mutate(Bump=c(rep(0,27),rep(1,21)))
nat_gas_clean$Month = 1:(dim(nat_gas_clean)[1])
nat_gas_clean = nat_gas_clean %>% mutate(MonthShort=c('Jan','Feb','March','April','May','June','July','Aug','Sep','Oct','Nov','Dec',
'Jan','Feb','March','April','May','June','July','Aug','Sep','Oct','Nov','Dec',
'Jan','Feb','March','April','May','June','July','Aug','Sep','Oct','Nov','Dec',
'Jan','Feb','March','April','May','June','July','Aug','Sep','Oct','Nov','Dec'))
#create data frame to store results, row names are the bldg codes and coefficients are the post-covid trend coefficient (Ramp)
nat_gas_results = data.frame(colnames(nat_gas_clean[1:175]),coefficient_value=0,row.names = colnames(nat_gas_clean[1:175]))
nat_gas_results = nat_gas_results[-1]
#colnames(elec_results)
nat_gas_results
for (i in colnames(nat_gas_clean)[1:175]) {
nat_gas_clean[i] <- as.numeric(unlist(nat_gas_clean[i]))
M_i = lm(unlist(nat_gas_clean[i]) ~ Ramp + Bump + Month + MonthShort,data=nat_gas_clean)
nat_gas_results[i,"coefficient_value"] = M_i$coefficients[3]
nat_gas_results[i,"mean"] = mean(unlist(nat_gas_clean[i]))
}
nat_gas_results
write.csv(nat_gas_results,"nat_gas_mean_trend.csv")
###############################
# Nat Gas Section End
###############################
################
# water section start
################
water = read_excel('UPC Utility Data 2018-2021.xlsx', sheet="Water", range = "A5:AY193") #this read cuts off the totals row which we don't happen to use and two houses without acronyms
water_clean = data.frame(t(water[c(-1),c(-1,-3)]))
head(water_clean)
dim(water_clean)
water_clean[1:187]
#janitor to get rid of auto generated column names and put proper column names on
water_clean = janitor::row_to_names(water_clean, row_number=1) %>% clean_names()
dim(water_clean)
#nat_gas_clean = nat_gas_clean[c(-176,-177,-178)]
#nat_gas_clean
water_clean = water_clean %>% mutate(Ramp=c(rep(0,27),1:21))
water_clean = water_clean %>% mutate(Bump=c(rep(0,27),rep(1,21)))
water_clean$Month = 1:(dim(water_clean)[1])
water_clean = water_clean %>% mutate(MonthShort=c('Jan','Feb','March','April','May','June','July','Aug','Sep','Oct','Nov','Dec',
'Jan','Feb','March','April','May','June','July','Aug','Sep','Oct','Nov','Dec',
'Jan','Feb','March','April','May','June','July','Aug','Sep','Oct','Nov','Dec',
'Jan','Feb','March','April','May','June','July','Aug','Sep','Oct','Nov','Dec'))
#create data frame to store results, row names are the bldg codes and coefficients are the post-covid trend coefficient (Ramp)
water_results = data.frame(colnames(water_clean[1:187]),coefficient_value=0,row.names = colnames(water_clean[1:187])) #187 columns to avoid selecting factor columns
water_results = water_results[-1]
#colnames(elec_results)
water_results
for (i in colnames(water_clean)[1:187]) {
water_clean[i] <- as.numeric(unlist(water_clean[i]))
M_i = lm(unlist(water_clean[i]) ~ Ramp + Bump + Month + MonthShort,data=water_clean)
water_results[i,"coefficient_value"] = M_i$coefficients[3]
water_results[i,"mean"] = mean(unlist(water_clean[i]))
}
water_results
write.csv(water_results,"water_mean_trend.csv")
###############################
# Water Section End
###############################
###############################
# Chilled Water Section Start
###############################
write.csv(water_results,"chill_water_mean_trend.csv")
###############################
# Chilled Water Section End
###############################