R function for finding beginning of a chr string-CodePudding

I am trying to figure out how to show the incidents happening only in 2016. The format is a chr format and looks like "2016-12-31". Is there a way to search for values only from 2016?

Current code:

most_fatalities_2016 <- gun_violence[which(gun_violence$date == "2016"), select = c("state", "city_or_county")]

I guess I'm looking for the r function that acts like the LIKE function in SQL. Any help?

CodePudding user response：

You can simply use substr() or grepl()

gun_violence[substr(gun_violence$date,1,4)=="2016",]

gun_violence[grepl("^2016-",gun_violence$date),]

The above returns all column of the gun_violence data.frame. If you want to only return specific columns, you can specify those columns like this:

gun_violence[grepl("^2016-",gun_violence$date),c("state", "city_or_county")]

CodePudding user response：

I may be going further than what is asked, but I want to give some advice regarding the way the data is being stored and manipulated.

It may be much easier in downstream analyses if we transform this character variable into a proper date format beforehand.

Advice #2: The dplyr package provides very clear synthax for manipulation of dataframes, which may be a nice introduction comming from SQL-based backgrounds.

Advice #3: Understanding and "reverse-engineering" the dbplyr package (https://dbplyr.tidyverse.org/) may be enlightening for SQL-experienced users

gun_violence$date <-readr::parse_date(gun_violence$date)

after that, we can use many date-related functions, such as:

library(dplyr)
library(lubridate)

gun_violence %>% filter(date < today())
###
gun_violence %>% filter(year(date) == 2016) ### for the desired operation in the question
###
and so on