Adding manually a legend element - ggplot2

Question

With the code below:

plt <- ggplot(data) + geom_step(aes(factor(no), var7, color = group_no, group = group_no), size = 1.6)
plt + geom_step(aes(factor(no), var5, color = group_no, group = group_no), linetype = 'dashed', size = .7)

I have generated the following plot.

Now I would like to modify legend in one of the following way:

• add the legend for color and dashed line
• add two dashed lines to the ones already visible on the legend and modify their explanation so that they would show for example 'id 1, var5', 'id2, var 5' for dashed lines and 'id1, var7', 'id2, var7' for solid lines (actually, this one is preferable)

Is it feasible? I have tried with scale_linetype_manual(values = c('var5', 'var7') but it didn't work.

The data look like this:

data <- structure(list(no = c(1L, 1L, 2L, 2L, 3L, 3L, 4L, 4L), group_no = structure(c(1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L), .Label = c("1", "2"), class = "factor"), var1 = c(2, 1, 1, 1, 1, 0, 1, 1), var2 = c(0, 0, 1, 1, 1, 
0, 1, 0), var3 = c(2.04466646181187, 0.660598114091747, 1.55142947390572, 
1.53055726052766, 1.33062973767801, 0.521466633696396, 0.383486796026974, 
0.320273289219046), var4 = c(0.786548055557462, 0.933132594335315, 
0.734844331310191, 0.404908113668656, 0.50963171017644, 0.066048513105941, 
0.156065948976073, 0.528480184907794), var5 = c(2, 1, 3, 
2, 4, 2, 5, 3), var6 = c(0, 0, 1, 1, 2, 1, 3, 1), var7 = c(2.04466646181187, 
0.660598114091747, 3.59609593571759, 2.19115537461941, 4.92672567339561, 
2.71262200831581, 5.31021246942258, 3.03289529753485), var8 = c(0.786548055557462, 
0.933132594335315, 1.52139238686765, 1.33804070800397, 2.03102409704409, 
1.40408922110991, 2.18709004602017, 1.93256940601771)), class = c("grouped_df", "tbl_df", "tbl", "data.frame"), row.names = c(NA, -8L), vars = "group_no", drop = TRUE, .Names = c("no", "group_no", "var1", "var2", "var3", "var4", "var5", "var6", "var7", "var8"), indices = list(c(0L, 2L, 4L, 6L), c(1L, 3L, 5L, 7L)), group_sizes = c(4L, 4L), biggest_group_size = 4L, labels = structure(list(group_no = structure(1:2, .Label = c("1", "2"), class = "factor")), class = "data.frame", row.names = c(NA, -2L), vars = "group_no", drop = TRUE, .Names = "group_no"))

Answer 1

If I understood correctly, here is an option for your first request

library(tidyverse)
data %>% 
 gather(variable, value, c(var5, var7)) %>% 
 ggplot() + 
 geom_step(aes(factor(no), 
               value, 
               color = group_no,
               linetype = variable,
               group = interaction(group_no, variable),
               size = variable)) +
 scale_linetype_manual(values = c('var5' = "dashed",
                                  'var7' = "solid")) +
 scale_size_manual(values = c('var5' = 1,
                              'var7' = 2)) +
 guides(size = "none")

Reshape your data from wide to long format and map color to 'group_no' and linetype to 'variable'. The interaction is necessary because the group is not defined by a single variable anymore, but by a combination of 'group_no' and 'variables'.

Change the size of the lines using scale_size_manual after you mapped 'variable' to this aesthetic. Add + guides(size = "none") such that the different sizes of geom_step is not displayed in the legend.