Visualizing Times Series
library(lubridate)##
## Attaching package: 'lubridate'## The following objects are masked from 'package:base':
##
## date, intersect, setdiff, unionlibrary(ggplot2)
library(scales)
library(gridExtra)
library(readr)##
## Attaching package: 'readr'## The following object is masked from 'package:scales':
##
## col_factorlibrary(tidyverse)## -- Attaching packages --------------------------------------- tidyverse 1.3.0 --## v tibble 3.0.5 v dplyr 1.0.3
## v tidyr 1.1.2 v stringr 1.4.0
## v purrr 0.3.4 v forcats 0.5.0## -- Conflicts ------------------------------------------ tidyverse_conflicts() --
## x lubridate::as.difftime() masks base::as.difftime()
## x readr::col_factor() masks scales::col_factor()
## x dplyr::combine() masks gridExtra::combine()
## x lubridate::date() masks base::date()
## x purrr::discard() masks scales::discard()
## x dplyr::filter() masks stats::filter()
## x lubridate::intersect() masks base::intersect()
## x dplyr::lag() masks stats::lag()
## x lubridate::setdiff() masks base::setdiff()
## x lubridate::union() masks base::union()harMetDaily.09.11 <- read_csv("Met_HARV_Daily_2009_2011.csv")## Warning: Missing column names filled in: 'X1' [1]##
## -- Column specification --------------------------------------------------------
## cols(
## .default = col_double(),
## date = col_date(format = ""),
## f.airt = col_character(),
## f.airtmax = col_character(),
## f.airtmin = col_character(),
## f.rh = col_character(),
## f.rhmax = col_character(),
## f.rhmin = col_character(),
## f.dewp = col_character(),
## f.dewpmax = col_character(),
## f.dewpmin = col_character(),
## f.prec = col_logical(),
## f.slrt = col_character(),
## f.part = col_character(),
## f.netr = col_character(),
## f.bar = col_character(),
## f.wspd = col_character(),
## f.wres = col_character(),
## f.wdir = col_character(),
## f.wdev = col_character(),
## f.gspd = col_character()
## # ... with 3 more columns
## )
## i Use `spec()` for the full column specifications.## Warning: 9 parsing failures.
## row col expected actual file
## 1087 f.prec 1/0/T/F/TRUE/FALSE E 'Met_HARV_Daily_2009_2011.csv'
## 1088 f.prec 1/0/T/F/TRUE/FALSE E 'Met_HARV_Daily_2009_2011.csv'
## 1089 f.prec 1/0/T/F/TRUE/FALSE E 'Met_HARV_Daily_2009_2011.csv'
## 1090 f.prec 1/0/T/F/TRUE/FALSE E 'Met_HARV_Daily_2009_2011.csv'
## 1091 f.prec 1/0/T/F/TRUE/FALSE E 'Met_HARV_Daily_2009_2011.csv'
## .... ...... .................. ...... ..............................
## See problems(...) for more details.# The data is available on the following website : https://data.neonscience.org/str(harMetDaily.09.11$date)## Date[1:1095], format: "2009-01-01" "2009-01-02" "2009-01-03" "2009-01-04" "2009-01-05" ...harMetDaily.09.11 %>%
select(date, airt) %>%
head(10)## # A tibble: 10 x 2
## date airt
## <date> <dbl>
## 1 2009-01-01 -15.1
## 2 2009-01-02 -9.1
## 3 2009-01-03 -5.5
## 4 2009-01-04 -6.4
## 5 2009-01-05 -2.4
## 6 2009-01-06 -4.9
## 7 2009-01-07 -2.6
## 8 2009-01-08 -3.2
## 9 2009-01-09 -9.9
## 10 2009-01-10 -11.1df <- harMetDaily.09.11 %>%
select(date, airt) ggplot(df, aes(date, airt)) +
geom_point(color = "purple",
size = 3,
pch = 18) +
labs(x= "Date",
y= expression(paste("Temperature ( ", degree ~ C, " )")),
title = "Air Temperature (2009-2011)\nNEON Harvard Forest Field Site") +
scale_x_date(labels = date_format("%b %y")) +
stat_smooth(colour = "green") +
theme(plot.title = element_text(color="darkblue", size=14, face="bold.italic"),
text = element_text(size=14,face="bold"),
axis.title.x = element_text(size=14, face="bold"),
axis.title.y = element_text(size=14, face="bold"))## `geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
# https://www.neonscience.org/resources/learning-hub/tutorials/dc-time-series-plot-ggplot-rAPPLICATION
1. Number of crimes in Toronto
Torontocrimes <- read_csv("Torontocrimes.csv")## Warning: Missing column names filled in: 'X1' [1]##
## -- Column specification --------------------------------------------------------
## cols(
## .default = col_double(),
## event_unique_id = col_character(),
## occurrencedate = col_datetime(format = ""),
## reporteddate = col_datetime(format = ""),
## premisetype = col_character(),
## offence = col_character(),
## reportedmonth = col_character(),
## reporteddayofweek = col_character(),
## occurrencemonth = col_character(),
## occurrencedayofweek = col_character(),
## MCI = col_character(),
## Division = col_character(),
## Neighbourhood = col_character()
## )
## i Use `spec()` for the full column specifications.nf <- count(Torontocrimes, occurrencedate)ggplot(nf, aes(occurrencedate, n)) + geom_line()
If we are interested in day
str(Torontocrimes$occurrencedate)## POSIXct[1:166500], format: "2014-06-20 10:55:00" "2014-07-02 00:20:00" "2014-07-02 00:20:00" ...Torontocrimes %>%
filter(occurrenceyear == 2018) %>%
select(occurrencedate) %>%
mutate(date = as.Date(occurrencedate)) %>%
group_by(date) %>%
summarise(frequency = n()) %>%
head(10)## # A tibble: 10 x 2
## date frequency
## <date> <int>
## 1 2018-01-01 177
## 2 2018-01-02 83
## 3 2018-01-03 67
## 4 2018-01-04 72
## 5 2018-01-05 100
## 6 2018-01-06 89
## 7 2018-01-07 90
## 8 2018-01-08 119
## 9 2018-01-09 87
## 10 2018-01-10 90Checking
Torontocrimes %>%
filter(occurrenceyear == 2018 & occurrencemonth == "January" & occurrenceday == 1) %>%
count()## # A tibble: 1 x 1
## n
## <int>
## 1 177nf <- Torontocrimes %>%
filter(occurrenceyear == 2018) %>%
select(occurrencedate) %>%
mutate(date = as.Date(occurrencedate)) %>%
group_by(date) %>%
summarise(frequency = n()) ggplot(nf, aes(date, frequency)) +
geom_line() +
theme(plot.title = element_text(color="darkblue", size=14, face="bold.italic"),
text = element_text(size=14,face="bold"),
axis.title.x = element_text(size=14, face="bold"),
axis.title.y = element_text(size=14, face="bold"))
Time-Series Calendar Heatmaps
Using ggTimeSeries package
library(ggTimeSeries)Torontocrimes %>%
select(occurrencedate) %>%
mutate(date = as.Date(occurrencedate)) %>%
group_by(date) %>%
summarise(frequency = n()) %>%
ggplot_calendar_heatmap("date", "frequency") +
xlab(NULL) +
ylab(NULL) +
facet_wrap(~Year, ncol = 1) +
scale_fill_gradient(high="red", low="yellow")+
theme(plot.title = element_text(color="darkblue", size=14, face="bold.italic"),
text = element_text(size=14,face="bold"),
axis.title.x = element_text(size=14, face="bold"),
axis.title.y = element_text(size=14, face="bold"))
library(gapminder)gapminder %>%
filter(country == "Canada") %>%
select(year, lifeExp) %>%
ggplot(aes(x=year, y=lifeExp)) +
geom_line()+
geom_point(size = 4, color = "blue")+
labs(x = NULL, y = NULL) +
theme(text = element_text(size = 16),
panel.border = element_rect(fill = NA, colour = "grey20")) +
labs(title = "Life expectancy in Canada",
x = NULL,
y = "Life expectancy (Years)") +
theme_minimal() +
theme(plot.title = element_text(color="darkblue", size=14, face="bold.italic"),
text = element_text(size=14,face="bold"),
axis.title.x = element_text(size=14, face="bold"),
axis.title.y = element_text(size=14, face="bold"))