Dygraphs

Basics

• dygraph() is the main function to call

• data : enter either time series data or data frame(numeric), for numeric data, the first column has to be the time.

• Range selector dyRangeSelector() : we can pipe this onto the original graph which let us select regions to look at

• dySeries() : display series and customize

• dyOptions() : specify color palette for series lines using colors option

      + `includeZero` 
      + `axisLineColor` 
      + `gridLineCOlor` 

• dyShading() : shading from chose time periods

• dyAxis() :drawFrid, fillGraph options

• time zones : dygraphs display time-series using time zone of the work machine. You can choose time zone withing the xts objset using useDataTimezone option

• Roll Periods dyRoller() : smooth out the display of a series, plotted point representing the average of the number of the time stamps in roll periods.

We will use ts datasets::EuStockMarkets dataset to demonstrate using dygraphs package

EuStockMarkets

Description

• daily closing prices of major stock indeces: DAX, SMI, CAC and FTSE

• 1860 observations and 4 variables

• mts object

• from 1991 - 1998. The data are sampled in business time, i.e., weekends and holidays are omitted.

ts<-datasets::EuStockMarkets
kable(cbind(head(time(ts), 6), head(ts, 6)), format="pandoc")

	DAX	SMI	CAC	FTSE
1991.496	1628.75	1678.1	1772.8	2443.6
1991.500	1613.63	1688.5	1750.5	2460.2
1991.504	1606.51	1678.6	1718.0	2448.2
1991.508	1621.04	1684.1	1708.1	2470.4
1991.512	1618.16	1686.6	1723.1	2484.7
1991.515	1610.61	1671.6	1714.3	2466.8

Exercise step by step

range selecting that allows zooming

#install.packages("dygraphs")
library(dygraphs);library(xts); 
dt<-data.frame(date=time(ts),as.matrix(ts))
dt$date=round(dt$date, 3)

## our data is sorted if not you need to sort by dates 
## load lubridate and xts 
## data$datetime= ymd_hms(data$datetime)
## then xts(x= data$Y, order.by=data$datetime)
## logscale == TRUE , specify in axis , ignored if x is date format 
dygraph(dt, main="European Stock Closing Price")%>%
dyRangeSelector()%>%dyLegend(show = "follow")%>%
  dyAxis("x",   axisLabelColor="black", axisLabelFontSize=1,   axisLineColor="red")

always show legends

dygraph(dt, main="European Stock Closing Price")%>%
dyRangeSelector()%>%dyLegend(show = "always")

Shading and Annotations + Vertical Shading

dygraph(dt, main="European Stocks")%>%
 #dySeries(label="DAX", color="black")%>% 
  dyLegend(show = "always", hideOnMouseOut = FALSE)%>%
  dyShading(from="1995", to="1996", color="#FFE6E6") %>%
  dyShading(from="1997", to="1998", color="#CCEBD6")

horizontal shading

mean<-apply(dt[,-1], 2, mean)
std<-apply(dt[,-1], 2, sd)
lower=mean-std
upper=mean+std

dygraph(dt, main="European Stocks")%>%
  dySeries(label="DAX", color="black") %>%
  dyShading(from=lower[1], to=upper[1], color="lightblack", axis = "y") 

Axis and Options

dygraph(dt, main = "EU stock") %>%
  dyAxis("y", label = "Stock Closing Price", valueRange = c(0, 10000)) %>%
  dyOptions(axisLineWidth = 1.5, 
            fillGraph = TRUE, 
            drawGrid = TRUE,
            colors = RColorBrewer::brewer.pal(4, "Dark2"),
            includeZero = TRUE, 
            axisLineColor = "navy", 
            gridLineColor = "lightblue")

Exercise step by step

before use useDataTimezone option

does not plot in the correct zone

##install.packages("xts")
library(xts)

ts.ran=seq.POSIXt(as.POSIXct("2019-01-01", tz="MST"),as.POSIXct("2019-07-02", tz="MST"), by="4 hours")

set.seed(123)
Y = rnorm(length(ts.ran))

## create time series object
serie= xts(Y, order.by = ts.ran, tz="MST")

dygraph(serie)%>%dyOptions(drawPoints = TRUE, pointSize = 2)%>%
dyRangeSelector()

after use useDataTimezone option

Correct!

dygraph(serie)%>%
dyRangeSelector()%>% dyOptions(useDataTimezone = TRUE, drawPoints = TRUE, pointSize = 2)

More

Roll Periods dyRoller()

In order to smooth out the series, we can use rollPerid = a number you choose. Add a rolling average period text box to the bottom left of the plot. Y values are averaged over the specified number of time scale units(e.g. days, months, years)

dygraph(serie, main = "simulated time series") %>%
dyRangeSelector()%>% dyOptions(useDataTimezone = TRUE, drawPoints = TRUE, pointSize = 2)%>% dyRoller(rollPeriod = 6)

stemPlot

dygraph(serie, main = "simulated time series") %>%
dyRangeSelector()%>% dyOptions(useDataTimezone = TRUE, drawPoints = TRUE, pointSize = 2)%>%dyOptions( stemPlot=TRUE)

stepPlot

dygraph(serie, main = "simulated time series") %>%
dyRangeSelector()%>% dyOptions(useDataTimezone = TRUE, drawPoints = TRUE, pointSize = 2)%>%dyOptions( stepPlot=TRUE)