Repository navigation
Expand file tree
/
Copy path0_load_data.R
More file actions
153 lines (104 loc) · 4.62 KB
/
Copy path0_load_data.R
File metadata and controls
153 lines (104 loc) · 4.62 KB
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
# Genome 2 M&V analysis
# written by Aoyu
#### SETUP ####
# use pacman
require(pacman)
# load libraries
pacman::p_load(tidyverse, lubridate, scales, slider, cowplot, patchwork, RColorBrewer, # general
broom, ggpmisc, ggpubr, # linear models
RMV2.0, # TOWT model
sprtt, effsize) # sequential testing
# turn off scientific notation
options(scipen = 999)
# set default theme for ggplot
theme_set(theme_minimal())
# define base ggplot theme
theme_update(plot.title = element_text(size = 14, colour = "grey20", face = "bold", hjust = 0.5),
plot.subtitle = element_text(size = 10, colour = "grey20", face = "italic", hjust = 0.5, margin = margin(b = 10)),
plot.caption = element_text(size = 8, colour = "grey20", face = "italic", hjust = 0.5),
plot.background = element_rect(fill = "white", colour = NA),
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
strip.text = element_text(size = 10, color = "grey20", face = "bold"),
strip.background = element_blank())
#### FUNCTION ####
perform_t_test <- function(df_pre, df_post, building) {
t.test(df_pre[[building]], df_post[[building]], alternative = "two.sided")
}
perform_mean_diff <- function(df_pre, df_post, building) {
abs(mean(df_pre[[building]], na.rm = T) - mean(df_post[[building]], na.rm = T)) / mean(df_pre[[building]], na.rm = T) * 100
}
perform_mean_est <- function(df_pre, df_post, building) {
0.5 * (sum(df_pre[[building]], na.rm = T) + sum(df_post[[building]], na.rm = T))
}
#### READ DATA ####
data_path <- "../../../Genome/buds-lab-building-data-genome-project-2/data/meters/cleaned/"
meta_path <- "../../../Genome/buds-lab-building-data-genome-project-2/data/metadata/"
weather_path <- "../../../Genome/buds-lab-building-data-genome-project-2/data/weather/"
output_path <- "../readfiles/"
df_elec <- read_csv(paste0(data_path, "electricity_cleaned.csv"))
df_meta <- read_csv(paste0(meta_path, "metadata.csv"))
df_weather <- read_csv(paste0(weather_path, "weather.csv"))
#### FILTER ####
# Focus
remove_type <- c()
remove_site <- c("Eagle", "Bobcat", "Swan", "Hog", "Gator", "Robin", "Lamb", "Moose", "Wolf", "Shrew", "Mouse", "Crow")
# NAs
na_counts <- sapply(df_elec[-1], function(x) sum(is.na(x)))
cols_to_keep <- names(na_counts[na_counts <= 1440])
cols_to_keep <- c("timestamp", cols_to_keep)
df_elec <- df_elec[, cols_to_keep]
# 0 means
means <- sapply(df_elec[-1], mean, na.rm = TRUE)
cols_to_keep <- names(means[means > 1])
cols_to_keep <- c("timestamp", cols_to_keep)
df_elec <- df_elec[, cols_to_keep]
# yearly difference
df_elec$timestamp = as.POSIXct(df_elec$timestamp, format="%Y-%m-%d %H:%M:%OS")
df_pre <- df_elec %>%
filter(year(timestamp) == 2016)
df_post <- df_elec %>%
filter(year(timestamp) == 2017)
buildings <- names(df_elec)[names(df_elec) != "timestamp"]
p_values <- sapply(buildings, function(building) {
t_test_result <- perform_t_test(df_pre, df_post, building)
t_test_result$p.value
})
mean_diff <- sapply(buildings, function(building) {
perform_mean_diff(df_pre, df_post, building)
})
mean_est <- sapply(buildings, function(building) {
perform_mean_est(df_pre, df_post, building)
})
# electricity intensity
eui_max <- 500
results <- data.frame(Building = buildings, P_Value = p_values, Dev = mean_diff, mean = mean_est) %>%
filter(Dev < 25) %>%
left_join(df_meta, by = c("Building" = "building_id")) %>%
mutate(elec_eui = mean / sqm) %>%
filter(elec_eui <= eui_max)
cols_to_keep <- c("timestamp", results %>% .$Building)
df_elec <- df_elec[, cols_to_keep]
# process corresponding metadata
df_meta <- df_meta %>%
filter(building_id %in% results$Building) %>%
select(building_id, sqm, sqft, eui) %>%
separate(building_id, into = c("site", "type", "name"), sep = "_") %>%
filter(!type %in% remove_type) %>%
filter(!site %in% remove_site)
# process corresponding weather data
df_weather <- df_weather %>%
select(timestamp,
site = site_id,
t_out = airTemperature) %>%
mutate(timestamp = as.POSIXct(timestamp, format="%Y-%m-%d %H:%M:%OS")) %>%
filter(site %in% df_meta$site)
#### OUTPUT ####
df_energy <- df_elec %>%
pivot_longer(c(-timestamp), names_to = "buildings", values_to = "eload") %>%
separate(buildings, into = c("site", "type", "name"), sep = "_") %>%
filter(!type %in% remove_type) %>%
filter(!site %in% remove_site)
write_rds(df_energy, paste0(output_path, "df_energy.rds"), compress = "gz")
write_rds(df_meta, paste0(output_path, "df_meta.rds"), compress = "gz")
write_rds(df_weather, paste0(output_path, "df_weather.rds"), compress = "gz")