Data Analysis: What Tools to Use?

As a data analyst, we use a combination of tools, including spreadsheet software, SQL databases, and Tableau. You can get quite a long way with these, but they’re not the only tools available. For each stage of the pipeline we’ll identify the tools we’re using alongside other popular and advanced alternatives you might encounter in your work.

Importing and tidying data

Spreadsheets programs such as Microsoft Excel, Google Sheets, or LibreOffice Calc provide a full set of features for importing and manipulating data. Tableau also provides some functionality for tidying raw data, but it lacks the flexibility of the spreadsheet applications.

In most cases, and for relatively small, simple data sets, the tidying process can be done manually. For large volumes of data, or those with more complex requirements, you can use programmatic solutions. Programming languages such as Python and R are popular in the data-science sphere and can be employed to ‘wrangle’ data. Both have built-in functions, or libraries, to speed up the process or perform common tasks.

Finally, if you are short of time, there are third-party tools and services designed specifically for wrangling data. These often claim to use AI or machine-learning technology to process your data, and suggest ways to structure it.

However you choose to tidy data, once it’s been prepared you can save it back into a common format (e.g. a spreadsheet or CSV file) or write it to a database before starting your exploratory analysis.

Exploratory data analysis

As you now know, exploratory data analysis has three stages: transform, visualise, and model. Initially, we’ll show you how to complete these stages with spreadsheet software, using a combination of its sorting and filtering capabilities and formulas to aggregate data during the transform stage. You’ll then use a spreadsheet’s chart features to create exploratory visualisations and fit simple models to your data.

More advanced tools (including R and Python) can transform data, and have additional packages for visualising data (e.g. ggplot2 in R, seaborn or matplotlib for Python).

Tableau also provides functionality for transforming, visualising, and applying basic models to data. You’ll learn about this in the second course.

To expand your skill set, you’ll also spend time learning about databases. In this course we’ll use the SQLite3 database. You’ll learn the SQL syntax to sort and filter data across multiple tables. There are many types of databases, and several other query languages or dialects. Popular relational databases include MySQL and PostgreSQL, both of which use SQL for querying or writing data.

Communication

Tableau is a popular business intelligence platform for creating visualisations and dashboards. Tableau can connect to a number of data sources – from spreadsheets and CSV files to database platforms – and uses intuitive drag-and-drop tools to create dashboards and charts. Beyond simple charts, Tableau provides advanced functionality to overlay patterns and trends. It also enables us to create highly interactive visualisations and to publish them for the public or others in your organisation. In the next course we’ll use Tableau to create charts and assemble a simple dashboard to present our data. To see examples of visualisations you can create with Tableau, check out the ‘viz of the day’ on the Tableau Public Gallery. [1]

Other tools, such as R Markdown or Jupyter notebooks, allow analysts to present their findings as reports, web pages, or technical documents that contain both code and visualisations. These tools are especially useful when presenting data to other analysts who are interested in your method as well as the results.

Reference

1. Gallery: viz of the day [Internet]. Tableau Public. Available from: https://public.tableau.com/en-gb/gallery/?tab=viz-of-the-day&type=viz-of-the-day