Online Statistical Computing Reference

• Time: August 2019 - Present
• Team contributors: Lizhou Fan, Kaixin Wang, Huizi Yu

Introduction

Online Statistical Computing Reference (OSCR) is an online statistical learning platform that students in social sciences could refer to. It is an online cookbook for learning data science in a non-statistical context.

Figure 1: Logo of OSCR

The main website of OSCR is publicly available at https://oscrproject.wixsite.com/website.

Our team currently consists of three students from the Department of Statistics at UCLA.

Languages

Languages that we are currently building are:

• main languages:

  • R
  • Python

• other languages frequently used in data analysis:

  • Excel
  • SQL
  • Tableau
  • Open Refine, Voyant Tools

Components

There are three main components in OSCR:

• Data Manipulation

  • data I/O
  • data cleaning and data preprocessing
  • data type coercion

• Data Modeling and Machine Learning

  • Supervised Learning

    • applications: prediction and classification

  • Unsupervised Learning

    • applications: clustering and PCA

• Data Visualization

  • in R: basic graphics, ggplot2, plotly, etc.
  • in Python: matplotlib, seaborn, etc.

Updates

As of Fall 2019, we are wrapping up the first section - Data Manipulation, and working on the Data Modeling and Machine Learning section.

Our schedule is for year 2019-2020 is as the following:

• finish Data Modeling and Machine Learning section by end of Fall 2019
• complete the section of Data Visualization by Winter 2020
• put the cookbook into practical use to help students with learning data science by end of Spring 2020