Human Mortality Database

This is an initial exploration of the Human Mortality Database, which is at https://www.mortality.org/.

Download the entire database and place it in your current working directory.

Setup

library(tidyverse)
## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr     1.1.4     ✔ readr     2.1.4
## ✔ forcats   1.0.0     ✔ stringr   1.5.1
## ✔ ggplot2   3.4.4     ✔ tibble    3.2.1
## ✔ lubridate 1.9.3     ✔ tidyr     1.3.0
## ✔ purrr     1.0.2     
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag()    masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors
library(plotly)
## Warning: package 'plotly' was built under R version 4.3.3
## 
## Attaching package: 'plotly'
## 
## The following object is masked from 'package:ggplot2':
## 
##     last_plot
## 
## The following object is masked from 'package:stats':
## 
##     filter
## 
## The following object is masked from 'package:graphics':
## 
##     layout

USAM Data

Load the data for USA males. Add a variable country and set it to “USA”.

Select country, Year, Age and qx.

Make Age numeric.

Eliminate any missing data.

Solution

USAM <- read_table("HMD/lt_male/mltper_1x1/USA.mltper_1x1.txt", skip = 2) %>% 
mutate(country = "USA") %>% 
select(country, Year, Age, qx) %>% 
mutate(Age = as.numeric(Age)) %>% 
filter(Age < 85) %>% 
rename(male_prob_death = qx) %>% 
drop_na()
## 
## ── Column specification ────────────────────────────────────────────────────────
## cols(
##   Year = col_double(),
##   Age = col_character(),
##   mx = col_double(),
##   qx = col_double(),
##   ax = col_double(),
##   lx = col_double(),
##   dx = col_double(),
##   Lx = col_double(),
##   Tx = col_double(),
##   ex = col_double()
## )
## Warning: There was 1 warning in `mutate()`.
## ℹ In argument: `Age = as.numeric(Age)`.
## Caused by warning:
## ! NAs introduced by coercion
summary(USAM)
##    country               Year           Age     male_prob_death   
##  Length:7650        Min.   :1933   Min.   : 0   Min.   :0.000100  
##  Class :character   1st Qu.:1955   1st Qu.:21   1st Qu.:0.001662  
##  Mode  :character   Median :1978   Median :42   Median :0.004540  
##                     Mean   :1978   Mean   :42   Mean   :0.019836  
##                     3rd Qu.:2000   3rd Qu.:63   3rd Qu.:0.024307  
##                     Max.   :2022   Max.   :84   Max.   :0.172840

Canada

Do the same for Canada.

Solution

CANM <- read_table("HMD/lt_male/mltper_1x1/CAN.mltper_1x1.txt", skip = 2) %>% 
mutate(country = "Canada") %>% 
select(country, Year, Age, qx) %>% 
mutate(Age = as.numeric(Age)) %>% 
filter(Age < 85) %>% 
rename(male_prob_death = qx) %>%
drop_na()
## 
## ── Column specification ────────────────────────────────────────────────────────
## cols(
##   Year = col_double(),
##   Age = col_character(),
##   mx = col_double(),
##   qx = col_double(),
##   ax = col_double(),
##   lx = col_double(),
##   dx = col_double(),
##   Lx = col_double(),
##   Tx = col_double(),
##   ex = col_double()
## )
## Warning: There was 1 warning in `mutate()`.
## ℹ In argument: `Age = as.numeric(Age)`.
## Caused by warning:
## ! NAs introduced by coercion
summary(CANM)
##    country               Year           Age     male_prob_death  
##  Length:8670        Min.   :1921   Min.   : 0   Min.   :0.00005  
##  Class :character   1st Qu.:1946   1st Qu.:21   1st Qu.:0.00141  
##  Mode  :character   Median :1972   Median :42   Median :0.00386  
##                     Mean   :1972   Mean   :42   Mean   :0.01870  
##                     3rd Qu.:1997   3rd Qu.:63   3rd Qu.:0.02082  
##                     Max.   :2022   Max.   :84   Max.   :0.18114

Combine

Combine the two dataframes into USA_CANM using rbind().

Solution

USA_CANM = rbind(USAM, CANM)

Male Infant Mortality USA and Canada

Produce a graph showing the probability of male death at age 0 for the USA and Canada. Use color to see two time-series plots. Create this graph beginning in 1940.

Solutiom

USA_CANM %>% 
  filter(Age == 0 & Year > 1940) %>% 
  ggplot(aes(x = Year, y = male_prob_death, color = country)) +
  geom_point() +
  ggtitle("Male Infant Mortality - USA and Canada")

USA/Canada 2

Create a graph comparing USA and Canadian male mortality at age 79.

Solution

USA_CANM %>% 
  filter(Age == 79 & Year > 1940) %>% 
  ggplot(aes(x = Year, y = male_prob_death, color = country)) +
  geom_point() +
  ggtitle("Age 79 Male Mortality - USA and Canada")

Task 1

Copy and modify the code above to produce USAF, CANF and USA_CANF. Do summaries to verify your work.

USAF <- read_table("HMD/lt_female/fltper_1x1/USA.fltper_1x1.txt", skip = 2) %>% 
mutate(country = "USA") %>% 
select(country, Year, Age, qx) %>% 
mutate(Age = as.numeric(Age)) %>% 
filter(Age < 85) %>% 
rename(female_prob_death = qx) %>% 
drop_na()
## 
## ── Column specification ────────────────────────────────────────────────────────
## cols(
##   Year = col_double(),
##   Age = col_character(),
##   mx = col_double(),
##   qx = col_double(),
##   ax = col_double(),
##   lx = col_double(),
##   dx = col_double(),
##   Lx = col_double(),
##   Tx = col_double(),
##   ex = col_double()
## )
## Warning: There was 1 warning in `mutate()`.
## ℹ In argument: `Age = as.numeric(Age)`.
## Caused by warning:
## ! NAs introduced by coercion
summary(USAF)
##    country               Year           Age     female_prob_death
##  Length:7650        Min.   :1933   Min.   : 0   Min.   :0.00008  
##  Class :character   1st Qu.:1955   1st Qu.:21   1st Qu.:0.00073  
##  Mode  :character   Median :1978   Median :42   Median :0.00297  
##                     Mean   :1978   Mean   :42   Mean   :0.01343  
##                     3rd Qu.:2000   3rd Qu.:63   3rd Qu.:0.01431  
##                     Max.   :2022   Max.   :84   Max.   :0.15084
CANF <- read_table("HMD/lt_female/fltper_1x1/CAN.fltper_1x1.txt", skip = 2) %>% 
mutate(country = "CAN") %>% 
select(country, Year, Age, qx) %>% 
mutate(Age = as.numeric(Age)) %>% 
filter(Age < 85) %>% 
rename(female_prob_death = qx) %>% 
drop_na()
## 
## ── Column specification ────────────────────────────────────────────────────────
## cols(
##   Year = col_double(),
##   Age = col_character(),
##   mx = col_double(),
##   qx = col_double(),
##   ax = col_double(),
##   lx = col_double(),
##   dx = col_double(),
##   Lx = col_double(),
##   Tx = col_double(),
##   ex = col_double()
## )
## Warning: There was 1 warning in `mutate()`.
## ℹ In argument: `Age = as.numeric(Age)`.
## Caused by warning:
## ! NAs introduced by coercion
summary(CANF)
##    country               Year           Age     female_prob_death 
##  Length:8670        Min.   :1921   Min.   : 0   Min.   :0.000030  
##  Class :character   1st Qu.:1946   1st Qu.:21   1st Qu.:0.000650  
##  Mode  :character   Median :1972   Median :42   Median :0.003125  
##                     Mean   :1972   Mean   :42   Mean   :0.013531  
##                     3rd Qu.:1997   3rd Qu.:63   3rd Qu.:0.013325  
##                     Max.   :2022   Max.   :84   Max.   :0.159520
USA_CANF <- rbind(USAF, CANF)
summary(USA_CANF)
##    country               Year           Age     female_prob_death
##  Length:16320       Min.   :1921   Min.   : 0   Min.   :0.00003  
##  Class :character   1st Qu.:1951   1st Qu.:21   1st Qu.:0.00069  
##  Mode  :character   Median :1974   Median :42   Median :0.00306  
##                     Mean   :1974   Mean   :42   Mean   :0.01348  
##                     3rd Qu.:1998   3rd Qu.:63   3rd Qu.:0.01371  
##                     Max.   :2022   Max.   :84   Max.   :0.15952

Task 2

Redo the graphs you produced above for females in the USA and Canada. Do you see the same patterns?

USA_CANF %>% 
  filter(Age == 0 & Year > 1940) %>% 
  ggplot(aes(x = Year, y = female_prob_death, color = country)) +
  geom_point() +
  ggtitle("Female Infant Mortality - USA and Canada")

USA_CANF %>% 
  filter(Age == 79 & Year > 1940) %>% 
  ggplot(aes(x = Year, y = female_prob_death, color = country)) +
  geom_point() +
  ggtitle("Age 79 Female Mortality - USA and Canada")

Task 3: Male + Female

Combine USAM and USAF into USA. This new dataframe will have both male and female probabilities of death. Run a summary to verify your work.

#Need to rename the male_prob_death and female_prob_death columns to make them the same or rbind will not work

USAM2 <- USAM %>% 
  rename(prob_death = male_prob_death) %>% 
  mutate(sex = "M")


USAF2 <- USAF %>% 
 rename(prob_death = female_prob_death) %>%
  mutate(sex = "F")


USA <- rbind(USAM2, USAF2)

Task 4: The Ratio

Compute a new variable ratio. It is the ratio of the male probability of death to the female probability. For the year 2019, plot this ratio with Age on the horizontal axis. Use geom_point().

ratio_m_f <- USA %>% 
  filter(Year == 2019) %>% 
  group_by(Age) %>% 
  summarize(ratio = prob_death[sex == "M"]/prob_death[sex == "F"])

ggplot(ratio_m_f, aes(x = Age, y = ratio)) +
  geom_point()

Task 5: Comments

Describe what you saw in Task 4. How would you explain this?

It is obvious that men have a much higher probability of death throughout their lives! It seems that once puberty kicks in, men really up their stupidity. This peaks in their mid 20s when the probability of death is approximately 2.6 times higher for men than women. Men start to wise up a little in their mid 20s and the ratio of probability of death starts to decline, with a steady decline to the early 40s. From here it levels out until the mid 60s where it starts to decline again. The best men can do when compared to women is at age 4, with a ratio of 1.15 compared to women. It is interesting that the ratio increases at ages 8 and again at 10, where a sharp spike is observed. Then the ratio actually declines for a few years. It is not to be though, because by age 14, men are off to the death races again!