poisson.Rmd

---
title: "Poisson regression"
csl: the-american-naturalist.csl
output:
  html_document:
    theme: cerulean
    toc: yes
  pdf_document:
    toc: yes
<!-- bibliography: references.bib -->
editor_options: 
  chunk_output_type: console
---

<!--
IMAGES:
Insert them with: ![alt text](image.png)
You can also resize them if needed: convert image.png -resize 50% image.png
If you want to center the image, go through HTML code:
<div style="text-align:center"><img src ="image.png"/></div>

REFERENCES:
For references: Put all the bibTeX references in the file "references.bib"
in the current folder and cite the references as @key or [@key] in the text.
Uncomment the bibliography field in the above header and put a "References"
title wherever you want to display the reference list.
-->

<style type="text/css">
.main-container {
  max-width: 1370px;
  margin-left: auto;
  margin-right: auto;
}
</style>

```{r general_options, include = FALSE}
knitr::knit_hooks$set(
  margin = function(before, options, envir) {
    if (before) par(mgp = c(1.5, .5, 0), bty = "n", plt = c(.105, .97, .13, .97))
    else NULL
  },
  prompt = function(before, options, envir) {
    options(prompt = if (options$engine %in% c("sh", "bash")) "$ " else "> ")
  })

knitr::opts_chunk$set(margin = TRUE, prompt = TRUE, comment = "",
                      collapse = TRUE, cache = FALSE, autodep = TRUE,
                      dev.args = list(pointsize = 11), fig.height = 3.5,
                      fig.width = 4.24725, fig.retina = 2, fig.align = "center")

options(width = 137)
```

Downloading the data:

```{r}
tmp <- tempfile(fileext = "xlsx")
download.file("https://www.dropbox.com/s/3uq5zaqt7tdqro8/Census%202015%20village%20dengue%20and%20built%20Up.xlsx?raw=1", tmp)
data <- readxl::read_excel(tmp)
```

Renaming the variables (replacing spaces with `_`)

```{r}
names(data) <- gsub(" +", "_", names(data))
```

```{r}
mod1 <- glm(All ~ as.factor(Urban_Type), poisson, data, offset = log(nb_person_2014))
summary(mod1)
anova(mod1, test = "LRT")
```

```{r}
mod2 <- glm(All ~ ordered(Born_in_this_place_CAT) + as.factor(Urban_Type), poisson, data, offset = log(nb_person_2014))
summary(mod2)
anova(mod2, test = "LRT")
cbind(exp(coef(mod2)), exp(confint(mod2)))
```

```{r}
mod3 <- glm(All ~ Moved_since_2005 + as.factor(Urban_Type), poisson, data, offset = log(nb_person_2014))
summary(mod3)
anova(mod3, test = "LRT")
cbind(exp(coef(mod3)), exp(confint(mod3)))
```

```{r}
with(data, plot(Moved_since_2005, log(All)))
with(data, plot(Moved_since_2005, log(All / nb_person_2014)))
with(data, plot(log(Moved_since_2005), log(All / nb_person_2014)))
```

```{r}
mod5 <- glm(All ~ Water_In_House + Moved_since_2005 + as.factor(Urban_Type), poisson, data, offset = log(nb_person_2014))
summary(mod5)
anova(mod5, test = "LRT")
cbind(exp(coef(mod5)), exp(confint(mod5)))
```

```{r}
mod6 <- glm(All ~ Water_In_House + as.factor(Urban_Type) + Moved_since_2005, poisson, data, offset = log(nb_person_2014))
summary(mod6)
anova(mod6, test = "LRT")
cbind(exp(coef(mod6)), exp(confint(mod6)))
```

```{r}
mod7 <- glm(All ~ as.factor(Urban_Type) + Work_Mig, poisson, data, offset = log(nb_person_2014))
summary(mod7)
anova(mod7, test = "LRT")
cbind(exp(coef(mod7)), exp(confint(mod7)))
```

```{r}
mod7 <- glm(All ~ Work_Mig + as.factor(Urban_Type), poisson, data, offset = log(nb_person_2014))
summary(mod7)
anova(mod7, test = "LRT")
cbind(exp(coef(mod7)), exp(confint(mod7)))
```

Final model, writing the residuals into a CSV file:

```{r}
mod7 <- glm(All ~ as.factor(Urban_Type) + Moved_since_2005, poisson, data, offset = log(nb_person_2014), na.action = na.exclude)
summary(mod7)
anova(mod7, test = "LRT")
cbind(exp(coef(mod7)), exp(confint(mod7)))
write.csv(data.frame(code_sig = data$code_sig, residuals = resid(mod7)), "residuals.csv", quote = FALSE, row.names = FALSE)
```