transform_and_merge_01.Rmd

---
title: "Share Example"
output: html_notebook
---

**Goal:** identify countries on opposite sides

## Disclaimers

- User should verify the results
- There is a potential to generate more concise code.  Meanwhile, this appears to work just fine.

## Load Libraries

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


## Example Data

```{r sampledata}
country <- c("A","B","C","D","A","E","F","X","Y","Z","Q","R")
dispute <- c(1,1,1,1,2,2,2,3,3,3,3,3)
side_a <- c(0,0,1,1,0,1,1,0,0,0,1,1)
side_b <- c(1,1,0,0,1,0,0,1,1,1,0,0)

country_df <- data_frame(country, dispute, side_a, side_b)

country_df

```

## Identify Sides 

### Side B Countries

```{r sidea}
side_a <- country_df %>% 
  filter(side_a == 0)
  
side_a
```

### Side A Countries

```{r sideb}
side_b <- country_df %>% 
  filter(side_a == 1)
  
side_b
```


## Transform Data ; Build Data Frame

```{r for_df}

# declare Tibble (data frame)
df_total = tibble()

for (i in 1:max(country_df$dispute)) {
  # Build vector of disputants (multi-valued cell: country_y) 
  country_y <- country_df %>% 
    filter(dispute == i) %>% 
    select(-dispute) %>% 
    mutate(side_b = if_else(side_b == 0, country, NA_character_)) %>% 
    select(side_b) %>% 
    drop_na() %>% 
    unique() %>% 
    as_vector() %>% 
    paste0(collapse = "|") 
  
  # Build a data frame that adds the disputants vector as a column-variable; 
  # drops blank rows ; separate the multi-valued cells (disputants/country_y)
  # into rows
  country_df_temp <- country_df %>% 
    filter(dispute == i) %>% 
    mutate(country_disputes = if_else(side_b == 1, country_y, NA_character_)) %>% 
    drop_na() %>% 
    separate_rows(country_disputes, convert = TRUE) 
  
  # Append big Dataframe from from the iterated dataframes
  df_total <- bind_rows(df_total, country_df_temp)
  
}

df_total%>% 
  arrange(country_disputes) %>% 
  rename(country_x = country,
         counrty_y = country_disputes) %>% 
  select(dispute, counrty_y, country_x)

```