Skip to content

Plotting

tito-kimbo edited this page Sep 10, 2018 · 3 revisions

Why bother?

Plotting provides visual representations of data and allows us to better understand relationships between features and even to find non-trivial correlations. Creative plotting and proper interpretation of the resulting graphs are key to understanding both Machine Learning and Data Science.

BMLF

In BMLF, we offer several plotting options, divided in a couple sections: basic plots and statistical plots.

Basic Plots

This section has the most basic and common graphical representations of data. Currently, it supports the following

  1. Bar plot: Represents data as vertical bars for magnitude comparison amongst other uses.
  2. Pie chart: Represents data as portions of a circle with size proportional to their magnitude.
  3. Line plot: Represents data as points and creates a line joining them in order to highlight trends.

Statistical Plots

This section contains statistics-specific representations of data and more complex visualizations. At the moment, available plots are

  1. Histogram: Representation of data repetition across a given dataset.
  2. Box plot: Informs of data point distribution, mean and variance; as well as allows further comparison with future measurements.
  3. Scatter plot: Represents data as points in a 2-dimensional space. Allows to visually notice groups, outliers and other trends.
  4. Histogram 2D: Same as histogram but with 2D data.

Clone this wiki locally