Why does one radar/spider plot had straight sides and another has curved edges, despite being created with the same data?

Question

I used the following geom_polygon code to create the radar plot below. In this figure, my four treatment (ES, EO, LO, LS) are at the four polygon corners, and my shapes are specific to 4 streams (Blue, Pink, Orange, Yellow). This code produced a four-sized polygon with straight edges. The same dataset is used to create both figures. Table here:

Substrate	Phase	vf
Cobble	ES	3.6
Pea Gravel	ES	4.5
Mix	ES	2.1
Sand	ES	1.9
Cobble	EO	5.6
Pea Gravel	EO	6.9
Mix	EO	3.2
Sand	EO	2.9
Cobble	LO	5.9
Pea Gravel	LO	7.4
Mix	LO	3.4
Sand	LO	3.1
Cobble	LS	1.8
Pea Gravel	LS	2.3
Mix	LS	1.1
Sand	LS	1.0

ggplot(data=nh420,  aes(x=phase, y=vf, group=substrate, colour=substrate))+ #
  annotate("text", x=4, y=c(2,4,6,8,10), label=c(2,4,6,8,10), hjust=0.5, vjust=1)+
  geom_polygon(size = 1.5, alpha= 0)+
  scale_y_continuous()+
  scale_x_discrete()+
  scale_colour_manual(values= c("#c55a11", "#ab3378", "#4677aa", "#cbbb44"))+
  #scale_fill_manual(values= c("#CA885F", "#B66C98","#4677aa","#cbbb44"))+
  #scale_y_continuous(expand=c(0,0),limits=c(0,10),breaks=seq(0,10,3))+
  #ylab(expression(NH["4"]~""^"+"-N~V["f"]~(m~d^{"-1"})))+
  theme_minimal()+
  theme(axis.title.x=element_blank(),
        axis.title.y=element_blank(),
        #axis.title.y=element_text(size=12,color="black"),
        #axis.text.y=element_text(size=10,color="black"),
        #axis.text.x=element_text(size=12,color="black"),
        axis.text.x=element_blank(),
        axis.text.y=element_blank())+
  theme(legend.position="none")+
  coord_polar(start = -9.5)

However, when I switch around the x variable and grouping to create the inverse figure, the sides of the polygon are now curved. Why the drastic change in shape? The code for the second image (gray and yellow lines) is below.

ggplot(data=nh420,  aes(x=substrate, y=vf, group=phase, colour=phase))+ #
  annotate("text", x=4, y=0:3, label=0:3, hjust=0.5, vjust=1)+
  geom_polygon(size = 1.5, alpha= 0)+
  scale_y_continuous()+
  scale_x_discrete()+
  scale_colour_manual(values= c("gray75", "yellow", "goldenrod", "gray35"))+
  #scale_fill_manual(values= c("#CA885F", "#B66C98","#4677aa","#cbbb44"))+
  #scale_y_continuous(expand=c(0,0),limits=c(0,10),breaks=seq(0,10,3))+
  #ylab(expression("Areal Uptake,"~italic(U)~(mg~N~m^{"-2"}~h^{"-1"})))+
  theme_minimal()+
  theme(axis.title.x=element_blank(),
        axis.title.y=element_blank(),
        #axis.title.y=element_text(size=12,color="black"),
        #axis.text.y=element_text(size=10,color="black"),
        #axis.text.x=element_text(size=12,color="black"),
        axis.text.x=element_blank(),
        axis.text.y=element_blank())+
  theme(legend.position="none")+
  coord_polar(start = -9.5)

I do not understand why the shape changes. Any help is appreciated!

Answer 1

I once had this problem but found a custom function/trick on Erwan Le Pennec blog

Step 1: Create a custom radar coordinate system:

coord_radar <- function (theta = "x", start = - pi / 2, direction = 1) {
  theta <- match.arg(theta, c("x", "y"))
  r <- if (theta == "x") "y" else "x"
  ggproto(
    "CordRadar", CoordPolar, theta = theta, r = r, start = start,
    direction = sign(direction),
    is_linear = function(coord) TRUE)
}

Step 2: Add the function to the plot object

ggplot(data=df,  aes(x=Substrate, y=vf, group = Phase, colour =Phase))+ 
  annotate("text", x=4, y=0:3, label=0:3, hjust=0.5, vjust=1)+
  geom_polygon(size = 1.5, alpha= 0)+
  scale_colour_manual(values= c("gray75", "yellow", "goldenrod", "gray35"))+
  scale_y_continuous()+
  scale_x_discrete()+
  theme_minimal()+
  coord_radar(start = -9.5) #NEW and Custom Coordinate System

Step 3: Optional, rearrange the levels to make it easier to read.

df$Substrate <- factor(df$Substrate,
                       levels = c(
                         "Mix",
                         "Pea Gravel",
                         "Cobble",
                         "Sand"
                       ))

Final Result, hope it helps.