R vs Python vs SAS
2 likes1,442 views
The document discusses the roles and characteristics of R, Python, and SAS in the context of business intelligence and big data. It highlights the importance of these programming languages for data analysis, predictive modeling, and machine learning, while emphasizing their ease of learning and demand in the job market. The author encourages learning any of these tools, particularly recommending R as a strong choice for aspiring analysts.
R vs Python vs SAS
- 1. R vs Python vs SAS Oliver Frost Wednesday, 18 January 2017 18/1/2017 Copyright Consolidata Ltd 2017 1
- 2. Today’s session: • A (very quick) introduction to business intelligence and the big data industry. • The role of the analyst. • What is R? What is Python? What is SAS? • Why should I learn them? • What can I use them for? 18/1/2017 Copyright Consolidata Ltd 2017 2
- 3. Oliver Frost GitHub: https://github.com/olfrost Twitter: @Consolidata LinkedIn: https://uk.linkedin.com/in/olliefrost Consolidata Ltd Twitter: @ConsolidataLtd http://www.consolidata.co.uk 18/1/2017 Copyright Consolidata Ltd 2017 3
- 4. Background • Cognitive Neuroscience BSc • Multiple disciplines – biology, chemistry, psychology, sociology: • Designing experiments • Data collection and research methods • Testing for significance, power calculations, predictive modelling • Data protection, data ethics • Now working as a data engineer: • Cleaning, reshaping and normalising survey data for a marketresearch company • Developing the ConsolidataData Platform. • Active member of the data analytics community 18/1/2017 Copyright Consolidata Ltd 2017 4
- 5. Working as an analyst • You may be familiar with some tools already, depending where you’ve come from: • Excel and Office tools • SPSS, MATLAB • SQL • BI and analytics are a bit of a continuous process: • Cleaning data – missing values? Bad data? • Reshape data – is the data in the right format? • Loading – how much is there? • Find patterns – do these patterns add value? • Presentation – can you tell a story? 18/1/2017 Copyright Consolidata Ltd 2017 5
- 6. What is R? • R is an open-source programming language, developed by academics and statisticians • Originally for maths and statistical analysis, but is slowly becoming an all-purpose language: • Collect and analyse social media data • Text analytics • Predict trends • Train machines to make predictions • Scrape data from websites • Also a great visualisation tool! 18/1/2017 Copyright Consolidata Ltd 2017 6
- 7. • It’s easy to learn • It’s free to use • R skills are in demand • The language is becoming increasingly popular • Open-source means you know exactly what your program is doing • Integration with other tools like Excel, SQL Server and pretty much any data analysis tool! • Shorter development cycles because new modules and packages are being released all the time What is R? 18/1/2017 Copyright Consolidata Ltd 2017 7
- 8. What is Python? • An all-purpose, general language that works on multiple platforms • High level and easy to learn like R • More commonly used for machine learning and predictive modelling (particularly good for academics and data scientists) • Open source and free to learn and use • More commonly by developers Source: http://spectrum.ieee.org/computing/software/the- 2016-top-programming-languages (IEEE - Institute of Electrical and Electronics Engineers) 18/1/2017 Copyright Consolidata Ltd 2017 8
- 9. What is SAS? • Statistical Analysis System • Stored data in tables and can be used for: • Writing reports • Developing applications • Data warehousing • Data mining • You don’t have to be technical… 18/1/2017 Copyright Consolidata Ltd 2017 9
- 10. What do businesses use these tools for? • Building “data pipelines”: • New data is coming in all the time • Needs to be extracted, transformed and loaded • Needs to be fast 18/1/2017 Copyright Consolidata Ltd 2017 10
- 11. What do businesses use these tools for? • Descriptive Analytics • These skills are in demand. • Businesses want to know about their historical data. • They also want to know what is happening right now. • New marketing opportunities? Save time and money in current processes? • Machine learning and data science? • Can our customers be divided into clusters? • Can we predict what a customer is likely to buy and make recommendations? • Can we detect fraud? Can we predict risk? 18/1/2017 Copyright Consolidata Ltd 2017 11
- 12. • Learning a language can be intimidating, especially from a non-technical background. • But from my experience, it was absolutely worth it. • No need to pick one tool over the other, they are all great. • I would recommend R, though… Conclusions 18/1/2017 Copyright Consolidata Ltd 2017 12