Key points of data mapping 3 - spaghetti map

Broken line diagrams with too many lines usually become unreadable. This kind of diagram is generally called spaghetti diagram. Therefore, this kind of chart can hardly provide information about the data.

Drawing example

Let's take the evolution of female baby names in the United States from 1880 to 2015 as an example.

# Libraries
library(tidyverse)
library(hrbrthemes)
library(kableExtra)
library(babynames)
library(viridis)
library(DT)
library(plotly)

# Display data
data <- babynames
head(data)
nrow(data)

A tibble: 6 × 5

year	sex	name	n	prop
<dbl>	<chr>	<chr>	<int>	<dbl>
1880	F	Mary	7065	0.07238359
1880	F	Anna	2604	0.02667896
1880	F	Emma	2003	0.02052149
1880	F	Elizabeth	1939	0.01986579
1880	F	Minnie	1746	0.01788843
1880	F	Margaret	1578	0.01616720

1924665

# Pick data for some names
data = filter(data,name %in% c("Mary","Emma", "Ida", "Ashley", "Amanda", "Jessica", "Patricia", "Linda", "Deborah",   "Dorothy", "Betty", "Helen"))
head(data)
nrow(data)

A tibble: 6 × 5

year	sex	name	n	prop
<dbl>	<chr>	<chr>	<int>	<dbl>
1880	F	Mary	7065	0.07238359
1880	F	Emma	2003	0.02052149
1880	F	Ida	1472	0.01508119
1880	F	Helen	636	0.00651606
1880	F	Amanda	241	0.00246914
1880	F	Betty	117	0.00119871

2599

# As long as women's data
data= filter(data,sex=="F")
head(data)
nrow(data)

A tibble: 6 × 5

year	sex	name	n	prop
<dbl>	<chr>	<chr>	<int>	<dbl>
1880	F	Mary	7065	0.07238359
1880	F	Emma	2003	0.02052149
1880	F	Ida	1472	0.01508119
1880	F	Helen	636	0.00651606
1880	F	Amanda	241	0.00246914
1880	F	Betty	117	0.00119871

1593

# mapping
ggplot(data,aes(x=year, y=n, group=name, color=name)) +
geom_line() +
scale_color_viridis(discrete = TRUE) +
theme(
  plot.title = element_text(size=14)
) +
ggtitle("A spaghetti chart of baby names popularity")

As can be seen from the figure, it is difficult to understand the evolution of the popularity of specific names according to one line. In addition, even if you try to follow a line to display the results, you need to associate it with a more difficult legend. Let's try to find some solutions to improve this graph.

Improvement method

Targeting specific groups

Suppose you draw many groups, but the actual reason is to explain the characteristics of a particular group compared with other groups. Then a good solution is to highlight the group: make it look different and give it an appropriate comment. Here, Amanda's popularity evolution is obvious. Keeping other names is important because it allows you to compare Amanda with all other names

# Add data item
data =  mutate( data, highlight=ifelse(name=="Amanda", "Amanda", "Other"))
head(data)

A tibble: 6 × 6

year	sex	name	n	prop	highlight
<dbl>	<chr>	<chr>	<int>	<dbl>	<chr>
1880	F	Mary	7065	0.07238359	Other
1880	F	Emma	2003	0.02052149	Other
1880	F	Ida	1472	0.01508119	Other
1880	F	Helen	636	0.00651606	Other
1880	F	Amanda	241	0.00246914	Amanda
1880	F	Betty	117	0.00119871	Other

ggplot(data,aes(x=year, y=n, group=name, color=highlight, size=highlight)) +
geom_line() +
scale_color_manual(values = c("#69b3a2", "lightgrey")) +
scale_size_manual(values=c(1.5,0.2)) +
theme(legend.position="none") +
ggtitle("Popularity of American names in the previous 30 years") +
geom_label( x=1990, y=55000, label="Amanda reached 3550\nbabies in 1970", size=4, color="#69b3a2") +
theme(,
  plot.title = element_text(size=14)
)

Use subgraph

Area maps can be used to provide a more comprehensive overview of the dataset, especially when used in conjunction with subgraphs. The evolution of any name can be easily glimpsed in the following chart:

ggplot(data,aes(x=year, y=n, group=name, fill=name)) +
geom_area() +
scale_fill_viridis(discrete = TRUE) +
theme(legend.position="none") +
ggtitle("Popularity of American names in the previous 30 years") +
theme(
  panel.spacing = unit(0.1, "lines"),
  strip.text.x = element_text(size = 8),
  plot.title = element_text(size=14)
) +
# Map by name
facet_wrap(~name)

As can be seen from the picture, Linda is a very popular name in a very short time. On the other hand, Ida has never been very popular and has been less used for decades.

Combination method

If you want to compare the evolution of each line with other lines, you can combine targeting specific groups and using subgraphs

# The duplicate column, name/name2, has different uses. One is used to display the data in the subgraph and the other is used to column
tmp <- data %>%
  mutate(name2=name)
head(tmp)

A tibble: 6 × 7

year	sex	name	n	prop	highlight	name2
<dbl>	<chr>	<chr>	<int>	<dbl>	<chr>	<chr>
1880	F	Mary	7065	0.07238359	Other	Mary
1880	F	Emma	2003	0.02052149	Other	Emma
1880	F	Ida	1472	0.01508119	Other	Ida
1880	F	Helen	636	0.00651606	Other	Helen
1880	F	Amanda	241	0.00246914	Amanda	Amanda
1880	F	Betty	117	0.00119871	Other	Betty

tmp %>%
ggplot( aes(x=year, y=n)) +
# Display data with name2
geom_line( data=tmp %>% dplyr::select(-name), aes(group=name2), color="grey", size=0.5, alpha=0.5) +
geom_line( aes(color=name), color="#69b3a2", size=1.2 )+
scale_color_viridis(discrete = TRUE) +
theme(
  legend.position="none",
  plot.title = element_text(size=14),
  panel.grid = element_blank()
) +
ggtitle("A spaghetti chart of baby names popularity") +
# Partition with name
facet_wrap(~name)

reference resources

• THE SPAGHETTI PLOT