Malcolm and I are working on changes in land use within the Municipality of Delta from 1990 to the present. We are examining two aspects: changes of land use within the ALR; and changes in agricultural/commercial/residential/land use. The work is supported by a grant from KPU and Environment Canada. Work so far:
The image below shows land use within agriculture over time
library(reshape)
cropyearstats <- read.csv("C:/Users/Stephen/Dropbox/delta/cropyearstats.csv")
cropslong <- melt(cropyearstats, id = "Year")
names(cropslong) <- c("Year", "Crop", "Ha")
library(ggplot2)
ggplot(cropslong, aes(x = Year, y = Ha, group = Crop)) + geom_smooth(aes(group = Crop,
color = Crop), se = F, size = 2)
The data to draw the plot was obtained by Malcolm who used GIS to label each parcel using data from Agri-Foods Canada and Environment Canada. He did this for the agricultural census years 1996, 2001, 2006 and 2011. The lines on the plot are the smoothed results of the area estimates. It is interesting to observe the decline in grass and the increase in berries.
I am also using this example for teaching……so here are my stats programming notes for those who are interested: the original dataset was in 'wide' format, looking like this:
cropyearstats
Year Berry Cash.Crop Grass Greenhouse Natural.Habitat Other.Agriculture
1 1996 305.1 3995 2597 18.84 201.6 121.8
2 2001 420.4 4126 1897 210.04 543.1 43.5
3 2007 864.7 3822 1974 178.53 519.4 178.1
4 2011 1210.6 3878 1565 204.25 556.1 158.7
Unknowns
1 242.42
2 273.04
3 88.31
4 184.11
We need to stack the data by year, so that it is in 'long' format. Hadley Wickham's 'reshape' package has the aptly named 'melt' function so that the melted data looks like this (after changing the names of the columns)
head(cropslong)
Year Crop Ha
1 1996 Berry 305.1
2 2001 Berry 420.4
3 2007 Berry 864.7
4 2011 Berry 1210.6
5 1996 Cash.Crop 3994.8
6 2001 Cash.Crop 4126.3
Then I used Hadley Wickham's ggplot graphics package to get the graph. I could have specified a colour for each crop and I think I'll do that when I have a bit more time. Otherwise ggplot just randomly assigns a colour. Fortunately here it gave grass a green colour but that is just luck.