using arrows

This post is on how to use arrows in a plot.

This post is to create a plot that has arrows indicating the direction

In the example, we use sample data, but if you know where you event starts and ends you can skip these steps and go directly to the final section.

Data

Load data from package SeaDens

library(SeaDens) #updated

This data is from a simulated survey.

survey_data<-survey_4326

Calculate gaps

Using the information from time, we will check when there were gaps and based on where a gap was, identify different events

Load tidyverse package to use some functions

library(tidyverse)

Check that your time data is in the correct format

survey_data$dt <- as.POSIXct(strptime(survey_data$timestamps, "%Y-%m-%d %H:%M:%S"))

This function uses the times to identify where there was a gap, assuming that the data is sorted

calculate_gaps<-function(my_data=my_data){
  time1<-my_data$dt
  time2<-lag(time1)
  time_dif<-as.numeric(difftime(time1,time2, units="mins"))
  my_data$time_dif<-as.numeric(time_dif)
  return(my_data)
}

After running the function, a new data frame will be created which includes a column named time_dif

survey_data_gaps<-calculate_gaps(survey_data)

Here we will define, how many minutes should be considered a gap

survey_data_gaps<-survey_data_gaps %>%
  mutate(gap_event = case_when(is.na(time_dif) ~ 'N',
                                time_dif >= 2 ~ 'Y',
                                TRUE ~ 'N'))

Identify events

To add a number to each event in order to be able to identify them separately, we use the following function

identify_events<-function(my_data=my_data){
  num_seq<-nrow(my_data)
  num_seq<-as.numeric(num_seq)
  my_data$num_seq<-as.numeric(paste(seq(1:num_seq)))
  subset_data<-subset(my_data,my_data$gap_event != "Y")
  subset_data$num_seq<-as.integer(subset_data$num_seq)
  subset_data$event_number<-(cumsum(c(1L, diff(subset_data$num_seq)) != 1L))
  subset_data$event_number<-subset_data$event_number+1
  subset_data$event_number<-stringr::str_pad(subset_data$event_number, 3, pad = "0")
  subset_data$event_number<-paste0("event_",subset_data$event_number)
  subset_data<-subset_data%>%select(num_seq,event_number)
  my_data_events<-full_join(my_data,subset_data,by='num_seq')
return(my_data_events)
}

The function will return a data frame with a new column called event_number

survey_data_events<-identify_events(my_data=survey_data_gaps)

Start and time of the events

Using the classification of the events, we will extract the first and the last location per event, which would become the start and end of the arrow on the plot

survey_time_events<-survey_data_events %>%
  group_by(event_number)%>%
  summarise(first_lat=first(latitude),
            last_lat=last(latitude),
            first_lon=first(longitude),
            last_lon=last(longitude))%>%
  drop_na()

Plot

To plot we use the function geom_segment and the additional argument arrow

ggplot(survey_time_events,
       aes(x = first_lon, y = first_lat)) +
  geom_segment(aes(xend = last_lon, yend = last_lat), arrow = arrow())

Re-scale

If you were need more frequent arrows, you can re-scale the events

Here I create a for loop to re-scale separately per event

rescale_events<-function(my_data=my_data,each_num=each_num){
  events_list<-split(my_data,my_data$event_number)
  new_events_list<-list()
  for( i in seq_along(events_list)){
    events_df<-events_list[[i]]
    rep_secuence<-rep(1:100, each=each_num)
    replicate_numbers<-rep_secuence[1:nrow(events_df)]
    events_df$rep_number<-replicate_numbers
    events_df$event_number2<-paste0(events_df$event_number,'-',events_df$rep_number)
    new_events_list[[i]]<-events_df
  }
new_events_df<- do.call("rbind",new_events_list)
return(new_events_df)
}

We define how often do we want the arrow to occur, here I selected 40, which represents the number of locations

survey_rescale<-rescale_events(my_data=survey_data_events,each_num=40)

Group by event

Use the function group_by and summarise to identify the start and end of the events, now with the rescaling there would be more events

survey_rescale_arrows<-survey_rescale %>%
  group_by(event_number2)%>%
  summarise(first_lat=first(latitude),
            last_lat=last(latitude),
            first_lon=first(longitude),
            last_lon=last(longitude))%>%
  drop_na()

Similarly to above we use the function geom_segment and the argument arrow

ggplot(survey_rescale_arrows,
       aes(x = first_lon, y = first_lat)) +
  geom_segment(aes(xend = last_lon, yend = last_lat), 
               arrow = arrow())

In a map

Finally, to see how it would look in a map we will include a base map from Germany.

Load the shapefiles and the package to plot

library(GermanNorthSea)
library(sf)

Transform to the corresponding CRS

German_land<-st_transform(GermanNorthSea::German_land, 4326)
German_EEZ<-st_transform(GermanNorthSea::German_EEZ, 4326)
German_coast<-st_transform(GermanNorthSea::German_coast, 4326)

Add the corresponding arguments and voilà

ggplot() + 
  geom_sf(data = German_EEZ, colour = 'black')+
  geom_sf(data = German_land, colour = 'black', fill = 'grey')+ 
  coord_sf(xlim = c(3, 9),ylim = c(53, 56))+
  theme_bw()+
  xlab('Longitude')+ylab('Latitude')+  
  geom_segment(data=survey_rescale_arrows,aes(x = first_lon, y = first_lat,xend = last_lon, yend = last_lat), 
               arrow = arrow(length=unit(0.10,"cm"), type = "closed"),
               color='#c1121f')

Further reading

To change the shape, size and form of the arrow visit geom_segment