Show and Tell R

You can use css to style your documents! (check out the Rmd code)

Using ## also works, so ## = h2

Cleaning JSON data

if you’ve ever come across json data, you’ll know how messy it can be and hard to use in R

this is where the jsonlite package can really come in handy

library(tidyverse) 
library(jsonlite)

messy_data <- tibble(
  subj = c("subj1", "subj2"),
  responses = c("{\"Q0\":\"Jenny\",\"Q1\":\"Student\",\"Q2\":\"27\",\"Q3\":\"Purple\"}",
                "{\"Q0\":\"Peter\",\"Q1\":\"Doctor\",\"Q2\":\"65\",\"Q3\":\"Blue\"}")
) # this is an example of what json data may look like

messy_data

## # A tibble: 2 x 2
##   subj  responses                                                              
##   <chr> <chr>                                                                  
## 1 subj1 "{\"Q0\":\"Jenny\",\"Q1\":\"Student\",\"Q2\":\"27\",\"Q3\":\"Purple\"}"
## 2 subj2 "{\"Q0\":\"Peter\",\"Q1\":\"Doctor\",\"Q2\":\"65\",\"Q3\":\"Blue\"}"

tidy_data <- messy_data %>%
  mutate(
    responses = map(responses, fromJSON), # from the jsonlite package
    name = map(responses, ~ unlist(.x)["Q0"]) %>% unlist(),
    occupation = map(responses, ~ unlist(.x)["Q1"])%>% unlist(),
    age = map(responses, ~ unlist(.x)["Q2"])%>% unlist(),
    color = map(responses, ~ unlist(.x)["Q3"])%>% unlist()
  ) %>%
  select(-responses)

tidy_data

## # A tibble: 2 x 5
##   subj  name  occupation age   color 
##   <chr> <chr> <chr>      <chr> <chr> 
## 1 subj1 Jenny Student    27    Purple
## 2 subj2 Peter Doctor     65    Blue