# SMHS Usage

## Contents

## Scientific Methods for Health Sciences - Learning and Instructional Usage

This Scientific Methods for Health Sciences EBook provides three types of instructional and learning materials:

- Datasets and natural driving motivational problems
- Mathematical techniques and modeling methodologies
- Applications with interactive graphical interfaces for statistical computing, data exploration and IT-blended instruction

Furthermore, all learning materials and instructional resources of this EBook are freely accessible via the Internet. This publication also includes multi-language support; at the bottom of any chapter, section or page, the reader may obtain a machine translation of all page content into a different language.

This EBook is not intended to be an *one-book-fits-all-curricula* textbook. Instructors that use the Scientific Methods for Health Sciences Ebook may customize some of the content, develop additional applications, discuss course-specific data, or even expand the materials as appropriate to their curricula.

The novelty of the Scientific Methods for Health Sciences EBook is derived from the fact that in the context of health sciences training, it integrates mathematical foundations, statistical concepts, interactive experiments, statistical computing resources, simulations, and tools for data analysis and visualization.

### SOCR Help

The SMHS Ebook uses SOCR Tools and R for all examples, computational demonstrations, data analytics and visualization. The following links provide useful tutorials and help with both software environments:

### R Statistical Computing Software

- Some R fundamentals
- The Software Carpentry Foundation provides useful Programming with R and R for Reproducible Scientific Analysis materials.
- A very gentle stats intro using R Book (Verzani)
- Quick-R examples
- R-tutor Introduction
- R project Intro
- UCLA ITS/IDRE R Resources

### Why Use R

There are marked differences between different types of computational environments for data wrangling, preprocessing, analytics, visualization and interpretation. The table below provides some rough comparisons between some of the most popular data computational platforms (*higher scores* represent better performance within the specific category, but the scales are not normalizes between categories).

More information about the rationale for using R is provided on the DSPA resources.

Language | OpenSource | Speed | ComputeTime | LibraryExtent | EaseOfEntry | Costs | Interoperability |
---|---|---|---|---|---|---|---|

Python | Yes | 16 | 62 | 80 | 85 | 10 | 90 |

Julia | Yes | 2941 | 0.34 | 100 | 30 | 10 | 90 |

R | Yes | 1 | 745 | 100 | 80 | 15 | 90 |

IDL | No | 67 | 14.77 | 50 | 88 | 100 | 20 |

Matlab | No | 147 | 6.8 | 75 | 95 | 100 | 20 |

Scala | Yes | 1428 | 0.7 | 50 | 30 | 20 | 40 |

C | Yes | 1818 | 0.55 | 100 | 30 | 10 | 99 |

Fortran | Yes | 1315 | 0.76 | 95 | 25 | 15 | 95 |

- NASA Comparison of Python, Julia, R, Matlab and IDL
- Sebastian Raschka, Numeric matrix manipulation - The cheat sheet for MATLAB, Python Nympy, R and Julia, June 2014.

- SOCR Home page: http://www.socr.umich.edu

Translate this page: