Julien Julien - 1 month ago 16
R Question

Scraping tables on multiple web pages with rvest in R

I am new to web scraping and am trying to scrape tables on multiple web pages. Here is the site: http://www.baseball-reference.com/teams/MIL/2016.shtml

I am able to scrape a table on one page rather easily using

rvest
. There are multiple tables, but I only wanted to scrape the first one, here is my code

library(rvest)
url4 <- "http://www.baseball-reference.com/teams/MIL/2016.shtml"

Brewers2016 <- url4 %>% read_html() %>%
html_nodes(xpath = '//*[@id="div_team_batting"]/table[1]') %>%
html_table()

Brewers2016 <- as.data.frame(Brewers2016)


The problem is that I want to scrape the first table on the page dating back to 1970. There is a link specifying the previous year at the top left corner just above the table. Does anybody know how I can do this?

I am also open to different ways of doing this, for example, a package other than rvest that might work better. I used rvest because it's the one I started learning.

Answer

One way would be to make vector of all the urls you are interested in and then use sapply:

library(rvest)

years <- 1970:2016
urls <- paste0("http://www.baseball-reference.com/teams/MIL/", years, ".shtml")
# head(urls)

get_table <- function(url) {
  url %>%
    read_html() %>%
    html_nodes(xpath = '//*[@id="div_team_batting"]/table[1]') %>% 
    html_table()
}

results <- sapply(urls, get_table)

results should be a list of 47 data.frame objects; each should be named with the url (i.e., year) they represent. That is, results[1] corresponds to 1970, and results[47] corresponds to 2016.