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     2.1.5
## ✔ forcats   1.0.0     ✔ stringr   1.5.1
## ✔ ggplot2   3.5.1     ✔ tibble    3.2.1
## ✔ lubridate 1.9.3     ✔ tidyr     1.3.1
## ✔ 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)
## 
## 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

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 = "Canada") %>% 
  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?

# Female Infant Mortality (Age 0) for USA and Canada
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 (Age 0)") +
  labs(y = "Probability of Death (Female, Age 0)")

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

# The patterns are very similar though the ffemale probablity of death at 79 is lower

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.

USA <- USAM %>%
  rename(male_prob_death = male_prob_death) %>%
  inner_join(USAF %>% rename(female_prob_death = female_prob_death), 
             by = c("Year", "Age", "country"))

# Display a summary of the combined dataframe to verify
summary(USA)
##    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  
##  female_prob_death
##  Min.   :0.00008  
##  1st Qu.:0.00073  
##  Median :0.00297  
##  Mean   :0.01343  
##  3rd Qu.:0.01431  
##  Max.   :0.15084

Task 4: The Ratio

# Compute the ratio of male to female probability of death
USA <- USA %>%
  mutate(ratio = male_prob_death / female_prob_death)

# Filter for the year 2019
USA_2019 <- USA %>%
  filter(Year == 2019)

# Plot the ratio with Age on the x-axis
ggplot(USA_2019, aes(x = Age, y = ratio)) +
  geom_point() +
  ggtitle("Ratio of Male to Female Probability of Death by Age (2019)") +
  xlab("Age") +
  ylab("Ratio of Male to Female Probability of Death")

## Task 5: Comments

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

I believe that men tend to live lives that exhibit higher levels of risk taking than females of the same age. These could include things like sports, military service, use of drugs or alcohol and risky driving. The differences appear to become more apparent at an approximate age of twelve and then increase sharply until the age of 22 or so where men are approximately three times as likely to die than females. After that, the mortality ratio begins to decrease sharply until the age of 42 or so. After that it is relatively level until 65 where it dips again until age 85 where the ratio is back to approximately 1.