forked from bambooforest/salos
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Phoible.Rmd
135 lines (75 loc) · 5.33 KB
/
Phoible.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
---
title: "Salos 2024 summer school: Working with PHOIBLE"
author: "Steven Moran and Alena Witzlack"
date: "(`r format(Sys.time(), '%d %B, %Y')`)"
output:
github_document:
toc: true
---
***
The libraries you will need.
```{r, message=FALSE}
library(tidyverse)
```
# Introduction
The [PHOIBLE database](https://phoible.org) is an online repository of phonological inventory data. Let's have a look at the website, where we can explore by various categories:
* [Inventories](https://phoible.org/inventories)
* [Languages](https://phoible.org/languages)
* [Segments](https://phoible.org/parameters)
***
But what if we want to explore and work with the data programmatically? What's the first thing we need to do?
***
We need to access the data. So, where is it?
One option is to download the data from the website:
* https://phoible.org/download
Data comes in different data formats (and different data types).
In linguistics, data are usually curated and we have data sets in different stages of development. For example, the PHOIBLE data are curated and maintained in a working format in a [GitHub](https://github.com) repository here:
* https://github.com/phoible/dev
This repository contains the raw data and code (scripts) to transform different data formats into a single spreadsheet:
* https://github.com/phoible/dev/blob/master/data/phoible.csv
This is a rather large spreadsheet, so we have made a smaller aggregated one here:
* https://github.com/bambooforest/salos/blob/main/data/phoible.csv
When a data table (aka spreadsheet) is smaller, GitHub will display it online for you:
* https://github.com/bambooforest/salos/blob/main/data/phoible.csv
***
How do we load the data into R/RStudio? You can simply use the variable name that you loaded the data into. `df` is a commonly used abbreviation for "data frame", i.e., the data structure used with in R for example for tabular data (more below).
```{r}
df <- read_csv('data/phoible.csv')
```
How do we have a quick look at it?
```{r}
df
```
How do we programmatically look at its structure, i.e., how do we know what it has in its columns and rows?
```{r}
str(df)
```
# Tabular data
In this course, we are mainly going to be dealing with data in [plain text](https://en.wikipedia.org/wiki/Plain_text) and structured data in rectangular format, also known as **tabular data**. Here are some descriptions:
* https://en.wikipedia.org/wiki/Table_(information)
* https://papl.cs.brown.edu/2016/intro-tabular-data.html
* https://www.w3.org/TR/tabular-data-model/
Tabular data **can be stored in many ways**, e.g.:
* [CSV](https://en.wikipedia.org/wiki/Comma-separated_values)
* [Excel sheets](https://en.wikipedia.org/wiki/Microsoft_Excel)
* [Google sheets](https://en.wikipedia.org/wiki/Google_Sheets)
* [Numbers](https://en.wikipedia.org/wiki/Numbers_(spreadsheet))
* [SQLite](https://en.wikipedia.org/wiki/SQLite)
* [JSON](https://en.wikipedia.org/wiki/JSON)
Which of these formats above are stored in plain text?
In a table data format, every column represents a particular variable (e.g., a person's height, number of of vowels) and each row/record corresponds to a given member of the data set in question (e.g. a person, in a language). Tabular data are inherently rectangular and cannot have "ragged rows". If any row is lacking information for a particular column, a missing value (`NA`) is stored in that cell.
For most people working with small amounts of data, the data table is the fundamental unit of organization because it is both a way of organizing data that can be processed by humans and machines. In practice, to enter, organize, modify, analyze, and store data in tabular form -- it is common for people to use spreadsheet applications. You are probably familiar for example with Excel spreadsheets.
Many statistical software packages use similar spreadsheets and many are able to import Excel spreadsheets. R is no different.
***
Tabular (or table) data has several properties. It consists of [rows and columns](https://en.wikipedia.org/wiki/Row_and_column_vectors) in the linear algebra sense, and [rows](https://en.wikipedia.org/wiki/Row_(database)) and [columns](https://en.wikipedia.org/wiki/Column_(database)) in the relational database sense.
[Columns](https://en.wikipedia.org/wiki/Column_(database)) in tabular data contain a set of data of a particular type and contain (typically) one value (data type -- see above) for each row in the table.
Each [row](https://en.wikipedia.org/wiki/Row_(database)) in the table contains an observation, in which each row represents a set of related data, i.e., every row has the same structure and each cell in each row should adhere to the column's specification (i.e., that data type of that column).
In sum:
* Every **column** represents a particular variable (height, weight, number of vowels).
* Each **row/record** corresponds to a given member of the data set in question (e.g. a person, a sentence, a language, a phoneme, a measurement).
* Every record shares the same set of variables, i.e., **every row has the same set of column headers**.
* Tabular data are **inherently rectangular** and cannot have "ragged rows".
* Each value (each cell) is **known as a datum**.
* If any row is lacking information for a particular column a **missing value (NA)** must be stored in that cell.
***