Problem with the legend with scale_color_discrete() function in R

Question

I've this data

tempi_k_random <- c(2.849072363496035E-6,4.884230846336759E-6,7.420560612989667E-6,9.320163959989317E-6,1.125275640739933E-5,1.3818609943264999E-5,1.5799511999414483E-5,1.79927469933985E-5,1.958275275649878E-5,2.1785199360099794E-5,2.4198109784553963E-5,4.54269327069883E-5,6.518881674784684E-5,8.52658233138289E-5,1.0445134492757438E-4,1.268194860165957E-4,1.4795886846906084E-4,1.7116724615770838E-4,1.9454743217185282E-4,2.095390879427764E-4,4.1366770908908134E-4,6.202640042883945E-4,8.379609645905531E-4,0.0010454638386975343,0.001188950746965453,0.0014212418300653593,0.0017239767156862745,0.0018250641025641023,0.002074198717948718,0.0040149999999999995,0.0059812500000000005,0.008516666666666665,0.011816666666666666,0.0189875,0.020087499999999998,0.0222125,0.016399999999999998,0.019649999999999997,0.0282,0.04525,0.0523,0.06235,0.0798,0.09570000000000001,0.10475,0.12225,0.1352)

tempi_k_cinque <- c(2.185171516292494E-6,3.688659557278267E-6,5.086848744974997E-6,6.550152577270055E-6,8.190041618674074E-6,9.569229015801014E-6,1.130849767469026E-5,1.2768555980732062E-5,1.4900814742468114E-5,1.5916628540808872E-5,1.7213466225813496E-5,3.214082728715961E-5,4.5726078618563776E-5,6.314850708691957E-5,8.436387207939846E-5,9.751538176472509E-5,1.0571751299724081E-4,1.1758304250994831E-4,1.3748931047676842E-4,1.480619766765244E-4,3.119003987090575E-4,4.3609348952131457E-4,6.194213313568152E-4,7.484261873392308E-4,8.79766911150425E-4,0.0010233251633986926,0.001126502976190476,0.0013468323343323345,0.001551022727272727,0.002741190476190476,0.004115000000000001,0.0052,0.0084125,0.0105125,0.01695,0.019,0.013525,0.016325,0.02135,0.03055,0.038,0.045149999999999996,0.05875,0.07305,0.0888,0.1039,0.09995)

tempi_ndieci <- c(2.375526248273769E-6,4.256132067044228E-6,6.042428977111871E-6,8.615698869627676E-6,9.95753073736447E-6,1.1776845215643598E-5,1.3660545425457787E-5,1.5924493501608036E-5,1.7668195352664178E-5,1.9221459566688867E-5,2.045119723829764E-5,3.811749152074419E-5,6.038099988218772E-5,7.483761391315114E-5,9.55648261718686E-5,1.1421521459312587E-4,1.353618000970425E-4,1.593224149232047E-4,1.664030090266135E-4,1.8585104947233018E-4,3.6222878593326967E-4,5.813731241132015E-4,7.396324821243827E-4,8.994984672166082E-4,0.0011234896715603237,0.001304954374293577,0.0013452083333333332,0.0016280505952380952,0.0018492628205128204,0.0037657738095238096,0.00477625,0.007389583333333334,0.010470833333333334,0.0153375,0.0194375,0.0166,0.0146,0.01645,0.022949999999999998,0.038,0.0464,0.05675,0.0742,0.08495,0.09765,0.10375,0.11015000000000001)

tempi_nmenoc <- c(2.28491345365468E-6,3.529909336343534E-6,5.025295275675399E-6,6.514100794323373E-6,8.058699140564791E-6,9.471434112359757E-6,1.0944375376030776E-5,1.2384409686795722E-5,1.4036495365797998E-5,1.5624169208365355E-5,1.6777502364941904E-5,3.14744111238494E-5,4.5192996614103015E-5,6.243421855589104E-5,7.603017835619405E-5,8.73392281204343E-5,1.0611253966143871E-4,1.2097295470045547E-4,1.3288902062298435E-4,1.4510677837849405E-4,2.898812731152251E-4,4.287646414347314E-4,5.860436388208565E-4,7.231695424141076E-4,8.790873015873016E-4,0.0010133180684554523,0.0010897767857142857,0.001376007326007326,0.0014679807692307692,0.002784523809523809,0.004365,0.004866666666666666,0.008370833333333334,0.0112625,0.0145375,0.016237500000000002,0.014225,0.016300000000000002,0.0238,0.032049999999999995,0.0437,0.04775,0.0581,0.0665,0.08595,0.0955,0.10055)


#Input da 100 a 10'000'100
lunghezzaInput <- c(100,200,300,400,500,600,700,800,900,1000,1100,2100,3100,4100,5100,6100,7100,8100,9100,10100,20100,30100,40100,50100,60100,70100,80100,90100,100100,200100,300100,400100,500100,600100,700100,800100,900100,1000100,2000100,3000100,4000100,5000100,6000100,7000100,8000100,9000100,10000100)

I've create this data frame

df <- data.frame(lunghezzaInput,tempi_k_random,tempi_k_cinque,tempi_ndieci,tempi_nmenoc)

and with ggplot2 library I've made this representation :

options(scipen=999)

gg <- ggplot(df,aes(x=lunghezzaInput, y=tempi_k_random, colour = "blue")) +
  scale_color_discrete("Tempi") +
  geom_smooth(method="loess", se=F) + 
  geom_smooth(aes(y=tempi_k_cinque,colour = "red"), method = "loess", se=F) +
  geom_smooth(aes(y=tempi_ndieci, colour = "#3DE61A"), method = "loess", se=F) +
  geom_smooth(aes(y=tempi_nmenoc, colour = "#FFD13B"), method = "loess", se=F) +
  labs(subtitle="Analisi dei tempi di Median of Medians", 
       y="Tempo (s)", 
       x="Grandezza Input", 
       title="Median of Medians")
  

plot(gg)

But in this image I want to change the legend element's name.

How can I call them

"K = Random"
"K = n - 5"
"k = n/10"
"k = 5" ??

Answer 1

This type of problems generaly has to do with reshaping the data. The format should be the long format and the data is in wide format. See this post on how to reshape the data from wide to long format.

First, create a named vector of color values.
Then reshape the data, pipe directly to ggplot and then use only one geom_smooth.

library(dplyr)
library(tidyr)
library(ggplot2)

tempi_names <- c("K = random", "K = n - 5", "k = n/10", "k = 5")
gg_colors <- c('#0000FF', '#FF0000', '#3DE61A', '#FFD13B')
gg_colors <- setNames(gg_colors, names(df)[-1])

df %>%
  pivot_longer(
    cols = starts_with('tempi'),
    names_to = 'Tempi',
    values_to = 'Value'
  ) %>% 
  ggplot(aes(x = lunghezzaInput, y = Value, colour = Tempi)) +
  geom_smooth(formula = y ~ x, method = 'loess', se = FALSE) +
  scale_colour_manual(name = "Tempi", 
                      breaks = names(df)[-1],
                      labels = tempi_names,
                      values = gg_colors) +
  labs(subtitle="Analisi dei tempi di Median of Medians", 
       y="Tempo (s)", 
       x="Grandezza Input", 
       title="Median of Medians")