FYP.Rmd

---
title: "Final Year Project"
author: "Forest Analytics Research Office"
output: html_document
---

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


### 1. Creation of a github account to manage R code base.
 - Go to www.github.com
 - ***Status:*** Completed

### 2. Commencement of a latex document to serve as Final Year Report
 - Create a master FYP file ( something like "myFYP.tex")
 - Create a master Bib file ( something like "myBib.tex")
 - "Bib" is short for bibliography, which deals with citations.
 - ***Status:*** Actioned and Reverted to in subsequent items


### 3. Compiling a bibtex file.
 - go to scholar.google.com
 - For relevant entry, click on quotation symbol (Second from left, where first symbol is a "star")
 - A dialogue box, labelled "Cite", should appear on screen
 - Go to the bottom of this dialogue box, and clink on "Bibtex"
 - A plain text html file should appear. Copy and Paste the contents of this into your Bib file.
 - ***Status:*** In Progress

### 4. Important LaTex Skills
 - Practice creating equations in Latex.
 - Fraction, subscripts, superscripts, greek letters, Summation Symbols
 - Importing graphics into a latex document ( \includegraphics{})
 - ***Status:*** Reverted to in subsequent items

### 5. Key Forestry Terms
 - Develop a familarity with key forestry metrics and terms
 - DBH, Basal Area, Stems, Logs, Plot
 - Review context of Stem Taper Equations - (this is a good review article: https://www.fs.fed.us/nrs/pubs/gtr/gtr-nrs-p-167papers/27-larsen_2016-CHFC.pdf)
 - Also - Kevin to provide Brian with the BFC Mensuration Book 
 - ***Status:*** In Progress

### 6. The lmfor R package
 - install the lmfor R package
 - Get the documentation for the R package on CRAN.
 - Pay special attention to the Growth models ( Naslund, Wykoff etc)
 - ***Status:*** In Progress

### 7. Useful Statistical Knowledge
 - RMSE, AIC
 - Residuals
 - Mean Absolute Error  
 - ***Status:*** In Progress
 
### 8. LiDAR
 - Develop a familiarity with some key concepts of LiDAR.
 - As a practice exercise for Latex, write a short paragraph on LiDAR as a latex document.
 - ***Status:*** Initiated. Reverted to in subsequent items.
 
 
### 9. Bibliography
 - incorporating a bibliography ( based on a bib. file)

### 10. Latex Improvements : Graphics and Captions
 - *(Successor of item 2)*
 - adding labels and captions to the graphics 
 - Simply including "\label{graphic1.jpg}" and "\caption{graphic1.jpg}" to the "includegraphics" part of the   latex code.

### 11. Latex Improvements : Equations
 - Incorporating equations into the latex code. 
 - For the sake of expediency, you should probably focus on some of the growth model equations listed here 
 (see pages 9 and 10)
 <pre><code>
 https://cran.r-project.org/web/packages/lmfor/lmfor.pdf
 </code></pre>
 
### 12. Imputing Heights with lmfor
 - With regards to fitting the model, I would suggest trying out the "imputeHeights()" function in 
 lmfor on the "trees" data set.
 - Try out the default model, which I believe to be the Naslund model.
 - After that, see if you can adjust it to other models described in the CRAN reference (i.e. Wykoff, etc)


### 13. Residual Analysis 
 - The "imputeHeights()"  function does not require you to split the data up, it will create new columns on the data frame. 
 - What is of interest is the residuals, i.e. difference between observed Heights, and heights predicted by the model. I propose that you compute the residuals and from that, the RMSE.

- For later, you can analyse the residuals, categorized by Species (SPP) , and also by Geographic location.

### 14. Developing Geographic Categorical Variables 

 - Adding Geographic categorical variables would require the county code to be clearly identified as a stand-alone variable.  
- This can be carried out with the "mutate" (from dplyr) and "substr".