Yield Data Check.Rmd

---
title: 'UNHCR DRS Livelihoods - Crop Yield Data Check '
output:
  pdf_document:
    latex_engine: xelatex
    fig_width: 3
    fig_height: 3
    fig_crop: no
  html_document:
    df_print: paged
header-includes:
- \usepackage{fontspec}
- \setmainfont{Arial}
---

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

```{r plot, echo=FALSE, include=FALSE }

library(tidyverse)

###### Store the link to the raw dataset hosts in github 
drs_df_raw_link <- "https://raw.githubusercontent.com/CartONG/R_training/main/df19.csv?token=ALMJAS5FZNUMPT2YXW2HM23A4B6Z2"


# download file again but forcing R to read all columns as character
df19 <- read_csv(drs_df_raw_link,
                 col_types = cols(.default = "c"))


### Create a function

# Identify crops by country
crop_check <- function(drs_df, base_end = c("Baseline", "Endilne"), Country_ = "All"){
  require(tidyverse)
  '
  1 - Filter by country & EL/BL
  
  2 - Extract vector of crops
    a - vector crop 1
    b - vector crop2
    c - join vectors
  3 - output unique values
  
  '
  base_end <- match.arg(base_end)
  
  if (Country_ == "All")
    temp <- drs_df %>% 
    filter(BE == base_end)
  
  else
    temp <- drs_df %>% 
    filter(BE == base_end & Country == Country_)
   
  c1 <- unique(temp$Crop1)
  
  c2 <- unique(temp$Crop2)
  
  c <- unique(c(c1, c2))
  c <- c[!is.na(c)]
  c <- c[c != 0]
    
  
  return(c)
    
}

# Graph problematic data entries
crop_yield <- function(df_drs, Crop, base_end = c("Baseline", "Endline"), Country = "All", Threshold = 100, Bin_width = 100){
  
  base_end <- match.arg(base_end)
  
  #crop 1
  if (Country == "All")
    t1 <- df_drs %>% 
    select(BE, Crop1, Country, Crop1KG, Crop1HA) %>% 
    filter(BE == base_end & Crop1 == Crop)
  
  else 
    t1 <- df_drs %>% 
      select(BE, Crop1, Country, Crop1KG, Crop1HA) %>% 
      filter(BE == base_end & Crop1 == Crop & Country == Country)
  
  t1 <- t1 %>% 
    rename(Crop = Crop1,
           CropKG = Crop1KG,
           CropHA = Crop1HA)
  

  #Crop2
  if (Country == "All")
    t2 <- df_drs %>% 
    select(BE, Crop2, Country, Crop2KG, Crop2HA) %>% 
    filter(BE == base_end & Crop2 == Crop)
  
  else 
    t2 <- df_drs %>% 
    select(BE, Crop2, Country, Crop2KG, Crop2HA) %>% 
    filter(BE == base_end & Crop2 == Crop & Country == Country)
  
  t2 <- t2 %>% 
    rename(Crop = Crop2,
           CropKG = Crop2KG,
           CropHA = Crop2HA)
  
  
  
  # combine crop lists
  t <- bind_rows(t1, t2)
  t$CropKG <- as.numeric(t$CropKG)
  t$CropHA <- as.numeric(t$CropHA)
  
  
  t <- t %>% 
    mutate(Yield = CropKG / CropHA,
           Check = Yield < Threshold)
  
  t_graph <- ggplot(t, aes(x=Yield)) +
    geom_histogram(data = subset(t, Check == TRUE), binwidth = Bin_width, fill = "red4")+
    geom_histogram(data = subset(t, Check == FALSE), binwidth = Bin_width, fill = "#0072BC")+
    labs(title = paste(Crop, "Yield (Kg/Ha) Per Seasson Per Sampled Beneficiary"))+
    xlab(paste(Crop,"yield (Kg/Ha) per season"))+
    ylab("# of sampled beneficairies")+
    scale_x_continuous(label=function(x){
      x <- x/1000
      return(paste(x, "To"))})
  
  return(t_graph)
    
  
}
```
## Burkina Faso
Crops:
```{r, echo=FALSE}

temp_x <- crop_check(df19, "Baseline", "Burkina Faso")
temp_x





```
## Plots
```{r pressure, echo=FALSE, fig.height=5,fig.width=5,out.height='50%',out.width='50%'}

for (c in 1: length(temp_x)){
  x <- temp_x[c]
  y <- crop_yield(df19, x, "Baseline", "Burkina Faso", Threshold = 200, Bin_width = 200)
  print(y)

}

```
.