ftips_meta.Rmd

---
title: "Analysis of Fingertips metadata"
author: "Julian Flowers"
date: "25/06/2017"
output: 
  ioslides_presentation: 
    incremental: yes
    keep_md: yes
    smaller: yes
    widescreen: yes
---

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

library(fingertipsR)
library(tidyverse)
library(stringr)
library(tidytext)
library(wordcloud)
library(topicmodels)
library(text2vec)
library(LDAvis)

```

## Metadata

```{r,  cache=TRUE}
prof_meta <- indicator_metadata(ProfileID = c(19, 40))

glimpse(prof_meta)

```

## Meta

```{r, cache=TRUE}
prof_meta %>% slice(7) %>%
  t() %>%
  data.frame() %>%
  knitr::kable()
prof_id <- indicators(ProfileID = c(19, 40)) 

prof_meta <- prof_meta %>%
  left_join(prof_id, by = c("Indicator.ID" = "IndicatorID"))

text <- prof_meta %>%
  select(Indicator.ID, DomainName, Indicator.full.name, Definition, Rationale)

```

## Comparison cloud
```{r, fig.height=5, fig.width=10}

text <- text %>%
  mutate_if(is.factor, as.character) %>%
  unite(text, c(Indicator.full.name, Definition, Rationale))

text$text <- tm::removeNumbers(text$text)
text$text <- tm::removePunctuation(text$text)
text$text <- tolower(text$text)

tidymeta <- text %>%
  unnest_tokens(word, text) %>%
  anti_join(stop_words) %>%
  arrange(DomainName) %>%
  count(DomainName, word, sort = TRUE) %>%
  reshape2::acast(word ~ DomainName, 
                  value.var = "n", fill = 0) %>%
  comparison.cloud(max.words = 500, title.size = .8, random.order = FALSE, rot.per = 0.4)

```





## Commonality cloud

```{r , echo = TRUE}

text %>%
  unnest_tokens(word, text) %>%
  anti_join(stop_words) %>%
  arrange(DomainName) %>%
  count(DomainName, word, sort = TRUE) %>%
  reshape2::acast(word ~ DomainName, 
                  value.var = "n", fill = 0) %>%
  commonality.cloud(max.words = 100)

```