0

[Edited Title and crossed off questions that were secondary to the main question]

What I want is a simple line plot with points (easy) with a background created by a data frame of tier levels (hard for me)

Example for what I am trying to recreate example plot CODE

# libraries used  
library(dplyr)  
library(ggvis)  
library(magrittr)  

# create example data frame
months <- c("oct", "nov", "dec", "jan", "feb", "march", "april", "may",
"june")
#months <- 1:9
tier3low <- (rep(150, 9))
tier3high <- c(157, 158, 162, 162, 166, 167, 169, 172, 172)
tier2high <- c(164, 166, 170, 171, 176, 178, 180, 182, 182)
tier1high <- rep(185, 9)
tier_range <- seq(from = 150, to = 185, by = 4)
scores <- c(156, 163, 162, 172, 173, 174, 175, 177, 183)
df <- data.frame(months, tier3low , tier3high, tier2high, tier1high,
tier_range, scores)
# make levels
levels(df$months) <- c("oct", "nov", "dec", "jan", "feb", "march", "april",
"may", "june")

# what I have tried
df %>%
  group_by(months) %>% 
  ggvis(x = ~months, y = ~tier_range) %>%
  layer_ribbons(y = ~tier3low,  y2 = ~tier3high, stroke := "red") %>% 
  layer_ribbons(y = ~tier3high, y2 = ~tier2high, stroke := "yellow") %>% 
  layer_ribbons(y = ~tier2high, y2 = ~tier1high, stroke := "green") %>% 
  layer_lines(y   = ~scores) %>% 
  layer_points(y  = ~scores)

Which produces this
enter image description here

Edited out these questions as per requests from comment 1. The plots are not filling in with each other 2. The actual line are not printed on the plot 3. The data has been rearranged and in the wrong months

MAIN QUESTION

The plot isn't filling between each other and I don't understand why this isn't the correct solution. What in your opinion would be a better way to accomplish this.

mtelesha
  • 2,079
  • 18
  • 16
  • Consider asking separate questions for each of the problems you are having. You may want to cook up smaller, easier to understand example data and graphs for each problem so that readers/answerers will have an easier time understanding what points are confusing you. – Curt F. Feb 17 '15 at 18:09
  • Well I posted what I wanted and what I tired BUT my main question is one question. How do you have a ribbon plot as a background. I thought I gave good example of what I did to get to where I was expecting. If I drop it to just the back ground I would expect that people would want more information. – mtelesha Feb 17 '15 at 18:24
  • You definitely have put in some work already to try to solve your problem. I appreciate that but right now the question is still like "debug my code" instead of like "I don't understand why Y happened when my code was X", i.e. about specific points of confusion. Possibly http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example would be a good resource to check if you haven't seen it already. – Curt F. Feb 17 '15 at 18:32
  • Getting voted down when I ask a question, show what I tried, show what i tried doesn't work and ask for a reason AKA different way to do what I did causes me to just stay away from ever asking. This and Hacker News have so changed over the years. I know ICQ I can jump on it and say my code doesn't work but this is something not seen in other questions. – mtelesha Feb 19 '15 at 13:39

1 Answers1

1

You actually weren't far away. There's three things you needed to fix:

You need to use fill, not stroke, to determine the color of the ribbons

The data frame needs to be sorted by month, as the ribbon draws from left to right

You had a surplus definition of y in your first call to ggvis, which I think confused things as you had to define y separately again in each layer (probably didn't do you harm but certainly confused me).

Here's what I think you need:

df %>%
   arrange(months) %>%
   ggvis(x = ~months) %>%
   layer_ribbons(y = ~tier3low,  y2 = ~tier3high, fill := "red") %>% 
   layer_ribbons(y = ~tier3high, y2 = ~tier2high, fill := "yellow") %>% 
   layer_ribbons(y = ~tier2high, y2 = ~tier1high, fill := "green") %>% 
   layer_lines(y   = ~scores) %>% 
   layer_points(y  = ~scores)

enter image description here

Peter Ellis
  • 5,694
  • 30
  • 46