Quantitative Data Management

class: center, top, title-slide

.title[
# Quantitative Data Management
]
.subtitle[
## CRMW Workshop 2
]
.author[
### Charles Lanfear
]
.date[
### 25 Jan 2024<br>Updated: 23 Jan 2024
]

---

# Today

A research question:

.text-center[
*Controlling for **density**, how is **deprivation** related to **crime** in London?*
]

<br>

![](slides_quantitative-data-management_files/figure-html/mod-dag-1.svg)

<br>

.pull-left[
Today we will:

* Create a basic project
* Load data we will need
* Prepare the data for analysis

]

.pull-right[
Next time we will:

* Visualize our data
* Model our outcomes
* Diagnose and (try to) address problems
]

---
# Setup

1. Open RStudio

2. In the project menu (top right) select *New Project...*

3. Select *New Directory*
   * Place it wherever you want

4. Using the files tab of the bottom-right panel
   * Make sure you are in your project's main directory
   * Create a new `code` folder
   * Create a new folder called `data`
   * Create a folders called `raw` and `derived` in `data`

5. Browse to this lecture on the course website
   * [clanfear.github.io/ioc_crmw/](clanfear.github.io/ioc_crmw/)

---
# The Data

Save these to the `data` directory in your project

* London Crime data - 2022

+ Crime outcomes
   + [Direct download link (.zip)](https://clanfear.github.io/ioc_crmw/_data/dpu-crime-met-2022.zip) — **multiple files!**
   + Source: [`data.police.uk`](https://data.police.uk/data/)

* LSOA Indices of Deprivation - 2019

+ Deprivation predictors
   + [Direct download link (.csv)](https://clanfear.github.io/ioc_crmw/_data/imd2019lsoa.csv)
   + Source: [`opendatacommunities.org`](https://opendatacommunities.org/resource?uri=http%3A%2F%2Fopendatacommunities.org%2Fdata%2Fsocietal-wellbeing%2Fimd2019%2Findices)

* LSOA Population Density - 2014

+ Density control
   + [Direct download link (.xlsx)](https://clanfear.github.io/ioc_crmw/_data/land-area-population-density-lsoa11-msoa11.xlsx) — a multi-sheet **Excel file!**
   + Source: [data.london.gov.uk](https://data.london.gov.uk/dataset/super-output-area-population-lsoa-msoa-london)

---

# Get to Work!

We want to produce **analysis-ready data**:

* Cross-sectional (one row per unit)
* Columns for predictors
* Columns for outcomes

The process:

* **Load and clean** up each file with *separate scripts*
* **Save derived data** as *separate files*
* **Join together** in another script and save the analysis data

We'll have *at least* **four scripts**!

* We'll start by making `1_process-metro.R`
   * Numbers make **run order** clear

.text-center[
*Let's work on this together!*
]

---
class: inverse

# Cleaning Data

<br>

![](img/owl.jpg)

---
class: inverse

# Wrap-Up

### Practice!

* Data management is the hardest and most time consuming part of any project
* You get good with **practice** and **intentional improvement**