5censusRegressionv2.Rmd

---
title: "5 Census Regressions: 2011 vs each decade, all and cities"
author: "Dan Olner"
date: "15 July 2017"
output:
  pdf_document:
    toc: true
    toc_depth: 2
---

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

```{r echo = F, warning = F, message = F}
#Preload libraries for myself
library(readr)
#for resizing images
#http://stackoverflow.com/questions/15625990/how-to-set-size-for-local-image-using-knitr-for-markdown
library(png)
library(grid)
library(dplyr)
library(ggplot2)
library(RColorBrewer)
library(scales)
library(tidyr)

options(scipen=10)

#source("Function_CoBRegressionFunctions.R")

sheet5 <- readRDS('R_data/sheet5.rds')

#Only four cobs
#sheet5sub <- readRDS('R_data/sheet5sub.rds')

#summary(lm(data = inputdata, formula = lmfunction)),

#lapply(split(inputdata,inputdata[,splitByFactor]), function(x) summary(lm(data=x, formula = lmfunction)))

#Separate city sheets
citiesList <- c('Glasgow','Edinburgh','Aberdeen','Dundee')

citySheets5 <- readRDS('R_data/citySheets5.rds')
#citySheets5sub <- readRDS('R_data/citySheets5sub.rds')
```
\clearpage



#2011 ~ 1971 to 2001: all CoBs, all zones

Formula (d = 1971 to 2001): *xij2011 ~ xij-d + w-dq + migshareMinusOwn-d + ea-d + popPerAcre-d*

```{r echo = F, comment=NA}

decades = c('1971','1981','1991','2001')

for(i in decades){
  
  print(paste0('DECADE 2011 ~ ',i))
  
  print(summary(lm(data=sheet5, formula = paste0('xij2011 ~ xij',i,' + w',i,'q + migshareMinusOwn',i,' + ea',i,' + popPerAcre',i))))
  
}

```


#Four cities (urban zones): 2011 ~ the other four decades

```{r echo = F, comment=NA}

for(j in 1:4){
  
  print(citiesList[[j]])
  
  for(i in decades){
  
    print(paste0('DECADE 2011 ~ ',i))
    
    print(summary(lm(data=citySheets5[[j]], formula = paste0('xij2011 ~ xij',i,' + w',i,'q + migshareMinusOwn',i,' + ea',i,' + popPerAcre',i))))
  
  }

}
```