Finding change over time from a table in R - fitting a model to the data

Question

this is the table that has the information, it is much larger but I have just included a small part of it

Part 1 : (solved) This might seem quite easy, but I am very stuck on creating code to make a graph which shows the change over time. The table shows infant mortality and I just need a graph showing how the infant mortality of one country has changed over time.

Part 2 : how can I fit a line to the data - for example, for just one country? I need to model the data as a linear response. How can I do this?


my_data <- structure(list(`1800` = c("Afghanistan", "Albania", "Algeria", 
"Andorra", "Angola", "Antigua and Barbuda", "Argentina", "Armenia", 
"Australia", "Austria"), `1801` = c(469L, 375L, 460L, NA, 486L, 
474L, 402L, 371L, 391L, 387L), `1802` = c(469L, 375L, 460L, NA, 
486L, 470L, 402L, 371L, 391L, 373L), `1803` = c(469L, 375L, 460L, 
NA, 486L, 466L, 402L, 370L, 391L, 359L), `1804` = c(469L, 375L, 
460L, NA, 486L, 462L, 402L, 370L, 391L, 346L), `1805` = c(469L, 
375L, 460L, NA, 486L, 458L, 402L, 369L, 391L, 333L), `1806` = c(469L, 
375L, 460L, NA, 486L, 455L, 402L, 369L, 391L, 321L), `1807` = c(470L, 
375L, 460L, NA, 486L, 451L, 402L, 368L, 391L, 309L), `1808` = c(470L, 
375L, 460L, NA, 486L, 447L, 402L, 368L, 391L, 325L), `1809` = c(470L, 
375L, 460L, NA, 486L, 444L, 402L, 367L, 391L, 316L)), row.names = c(NA, 
10L), class = "data.frame")

Perhaps you need a `matplot` `matplot(t('row.names<-'(as.matrix(my_data[-1]), my_data[[1]])), type = 'l')` — akrun, Aug 30 '20 at 19:13
What is the time variable? The columns' names are 1800, 1801, ..., 1809. The data doesn't seem to have a time variable. — Rui Barradas, Aug 30 '20 at 19:29
Those are years, I don't know exactly how to add a time variable but there are new values every year etc — user783029, Aug 30 '20 at 19:33
I’m voting to close this question because it is not a question — ThomasIsCoding, Sep 03 '20 at 11:29

score 0 · Answer 1 · answered Aug 30 '20 at 19:40

0

You can plot what the question asks for with package ggplot2.
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 long to wide format.

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

my_data %>% 
  rename(country = `1800`) %>%
  pivot_longer(
    cols = starts_with('18'),
    names_to = 'time',
    values_to = 'mortality'
  ) %>%
  mutate(time = as.numeric(time)) %>%
  ggplot(aes(time, mortality, color = country)) +
  geom_point() +
  geom_line()

answered Aug 30 '20 at 19:40

Rui Barradas

70,273
8
34
66

it keeps saying that it cannot find function pivot_longer - is there a reason for this? Thanks for your help – user783029 Aug 30 '20 at 21:12
@000elizah000 `pivot_longer` is a new function in package `tidyr`, introduced in version `1.0.0`. See what `packageVersion('tidyr')` gives, maybe it's time to update. – Rui Barradas Aug 30 '20 at 21:27

akrun · Answer 2 · 2020-08-30T19:54:23.303

0

We can use matplot from base R

matplot(t('row.names<-'(as.matrix(my_data[-1]), my_data[[1]])), 
          type = 'l', xaxt = 'n')
legend("top", my_data[[1]], col = seq_along(my_data[[1]]),
          fill = seq_along(my_data[[1]]))

edited Aug 30 '20 at 19:54

answered Aug 30 '20 at 19:44

akrun

874,273
37
540
662

Finding change over time from a table in R - fitting a model to the data

2 Answers2