Limit axes so that they go from one range to another without problem

Question

Limit axes so that they go from one range to another without problem

Navigation

#1 by (2 votes)

1

I am trying to generate a ternary graph with types of soils made up of different combinations of Arcilla , Arena and MO . In turn, I calculate the growth of the plants in the types of conformed soils. Values range from 0 to 100% in Arcilla and Arena , while for MO is from 0 to 10%. The problem is that using coord_tern and / or tern_limits I can not shorten the eje R (MO) so that the points are distributed throughout the graph and not only concentrate on one end since the scales default for the axes are from 0 to 100.

If I was not clear in the query, please let me know.

ggtern(data = Libro1,mapping = aes(x= Arcilla, y= Arena, z= MO)) +
coord_tern(Tlim = ("0,100"), Llim = ("0,100"), Rlim = ("0,10")) +
tern_limit(T = 1, L = 1, R= 0.1) + stat_density_tern(geom='polygon', 
  aes(fill=..level..), bins=5, color='grey') + geom_point()

r rstudio ggplot2

asked by Ignacio 12.09.2017 в 17:00

source

1 answer

ASP.NET execute method through AJAX Android Action_DIAL error "No Activity found to handle Intent {act = android.intent.action.DIAL"

score 2 · Accepted Answer

It is not possible to do what you need due to limitations of the ggtern package. Each axis in the graph must add the same as the rest. As T 0: 100, L 20: 120 and R 100: 200. In the case of what you want to do, raise the following error (for the same reason I explain):

library(ggtern)

Libro1 <- data.frame(Arcilla=1:100,Arena=1:100, MO = seq(from=0,to=10, length.out = 100))

ggtern(data = Libro1,mapping = aes(x= Arcilla, y= Arena, z= MO)) +
  scale_T_continuous(limits = c(0,100))+
  scale_L_continuous(limits = c(0,100))+
  scale_R_continuous(limits = c(0,10))

##     T   L  R Sum
## 1 100   0  0 100
## 2   0 100  0 100
## 3   0   0 10  10
## Error: Invalid Ternary Limits, Each Point Must Sum to Unity...

I recommend expressing the units differently (as MO x 10 ^ -1) to be able to use the same scale as Arena and Arcilla :

library(ggtern)

Libro1 <- data.frame(Arcilla=1:100,Arena=1:100, MO = seq(from=0,to=10, length.out = 100))

Libro1$MO <- Libro1$MO*10

ggtern(data = Libro1,mapping = aes(x= Arcilla, y= Arena, z= MO)) +
  Tlab('Arena')+Llab('Arcilla')+Rlab(expression(paste('MOx10'^-1)))