Data Science

Mix of projects showcasing some of data science work

A report is available for some of them, hosted at pedrosan.github.io

• Rent vs. Buy (simulation-based) Calculator

  • The Calculator on shinyapps.io
  • Introduction and general information
  • Analysis of stock market returns (indices) to get a handle on how to include them in the calculator.
  • Repository

• myTextPredictr, a text prediction application

  • The Application on shinyapps.io
  • Progress Report
  • Final Report (in progress)
  • Slide deck with summary and basic information about algorithm
  • Repository

• Bike Sharing Data:

  • REPORT
  • Repository

• Baby Names Data:

  • REPORT
  • Repository

• Synthetic Data Set (challenge):

  • REPORT
  • Repository

• Analysis of Human Activity Data: Steps History and Patterns

  • REPORT
  • Repository

• ML-based Assessment of the Quality of Weight-lifting Exercises:

  • REPORT
  • Repository

• Human and Economic Cost of Major Storm Events: An Analysis of the NOAA/NCDC Database:

  • REPORT
  • Repository

Note on the Reports' Reproducible Format

The reports are generated from Rmarkdown documents that include all code to perform data processing, modeling, plotting, etc. (except for functions defined in external scripts, included in the repositories).

• For readability not all chunks of code are echoed explicitly in the compiled documents.
• In some cases the full straight reproducibility is limited by the fact that in the interest of simplicity and for computational convenience some parts of the processing have been flagged as inactive (eval = FALSE), and in some data are instead loaded from previously saved work (e.g. the model fitting). The document however include the code to perform the entire analysis.