---
title: "StringrExamples"
author: "Kathleen"
date: "September 21, 2017"
output: pdf_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

## R Markdown

```{r}
library(tidyverse)
library(stringr)

string1 <- "This is a string"
string2 <- 'If I want to include a "quote" inside a string, I use single quotes'
x <- c("\"", "\\")
x

x <- c("apple", "banana", "pear") 
# this is a character vector
str_length(x) # returns a vector with the length of each string
str_length("Welcome to DS4100")
str_length(c("kathleen", "gregg", "ava"))


str_c("k","c")
str_c("k","c", sep = "|")
str_c("k","c", sep = "|")
y <- c( "doll", "yoyo", NA)
str_length(y)
str_replace_na(y) # substitute the "NA" string for NA
str_c(c("k","t","d"),collapse ="") # make it 1 string
str_c(c("k","t","d"),collapse =",") # make it 1 string

#collapse reduces all elements to 1
str_c(c("k","t","d"),c("a","m","s"), collapse ="|") 

# str_c is vectorized 
str_c(c("k","t","d"),c("a","m","s"), sep =",") 
# result has 3 elements

#collapse reduces all elements to 1
str_c(c("k","t","d"),c("a","m","s"), sep = ",", collapse ="|" ) 

```

```{r}
colours <- c("red", "orange", "yellow", "green", "blue", "purple")
colour_match <- str_c(colours, collapse = "|")
colour_match

#built in sentences to use for testing
localSentences <- sentences

localSentences <- c("17 Barbara Rd", "18 Stella Rd.", "UNKNOWN")

# pull out substrings by position
str_sub("Apple",1,3)
str_sub(c("Apple","Pear", "Plum"), 1, 3)
str_sub(c("Apple","Pear", "Plum","Te"), 1, 3) # only 2 characters in last entry

#functions for cleaning up strings
str_to_lower("Apple")
str_to_upper("apple pear")
str_to_title("THIS IS NOT WHAT i WANT")
```
```{r}
str_trim("   Now is the time    ")
str_wrap(c("the time is notw right for revolting"), width = 1) 
# str_view takes a character vector and a regular expression and shows you the match in the string for the regular expression 

# an array of character vectors 
zz <- c("apple", "pear", "plum")

#find specific substrings
str_view(zz,"pl")
str_view(zz,"p")

# match 0 or 1 p's in the strings
str_view(c("apple","pear","plum"),"p?")

x <- c("apple", "banana", "pear") 
# this is a character vector
xp <- c("apple.", "banana.", "pear.") 
# this is another character vector with an embedded dot
xb <- c("apple\\", "banana\\", "pear\\") 
# this is a character vector with an embedded \

str_view(x, "an")
str_view(x, ".a.") 
# a dot matches any character but a new line 
str_view(xp, "e\\.") 
# match e followed by a dot 
str_view(xb, "r\\\\") 
# match r followed by a \ 
```


```{r }

# anchoring 
c("apple\\", "banana\\", "pear\\") %>% str_match("^a") # only match
c("apple\\", "banana\\", "pear\\") %>% str_view("^a") # only match
str_view("12345", "^\\d\\d\\d")
str_view("12345", "[^12]") # this means not 1 or 2 
str_view_all("12345", "[^12]")
str_view(x,"p?") # want 0 or 1 p in the string 
str_view(x,"p{1,2}")
palidrome <- c("anna", "eve", "hannah")
str_view(palidrome, "(.)(.)\\2\\1")
str_view(palidrome, "(.)(.)\\2\\1?")
str_view(palidrome, "(.)(.)(.)\\3\\2\\1?")

# detect if a string matches a regular expression
str_detect(x,"e")
sum(str_detect(x,"e"))
mean(str_detect(x,"e"))


```

```{r}
str_extract(x,"p..") #substrings that start with a p and are 3 chars.
xc <- x 
xd <- x
str_replace(xc, "p..", "x")
str_replace_all(xc, c("p"= "x", "l" = "y"))
str_split("apples corn pears", " ")
str_split("apples corn pears", "p..")
str_locate("apples corn pears", "p..")
str_locate(x, "p..") # returns a matrix
# matching substrings 
str_match(localSentences, "[1234567890]+")
str_match(c("adad","nana","anna"), "(..)+\\1")
str_match(c("adad","nana","anna"), "(..)+")
str_match(c("adad","nana","anna"), "(.)(.)+\\2\\1")
str_match(c("adada","nana","annaanna"), "(.)(.)\\2\\1")
str_detect(c("adada","nana","annaanna","adda", "addda"), "(.)(.)\\2\\2\\1")
str_match(localSentences, "\\d+")
ds <- c (localSentences, "123", "456")
str_replace(ds, "[1|2|3|4|5|6|7|8|9|0]", "-")
str_replace_all(ds, "[1|2|3|4|5|6|7|8|9|0]", "-")
```

```{r}
library(forcats)
month_levels = c( "Jan", "Feb", "Mar", "Apr", "May", 
                  "Jun", "Jul", "Aug", "Sep", "Oct", "Nov", "Dec")
x1 <- c("Dec", "Apr", "Jan", "Mar")
y1 <- factor(x1, levels = month_levels)
y1
#> [1] Dec Apr Jan Mar
#> Levels: Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
sort(y1)
```


```