Scripts to help Exploratory Data Analysis
# MagikEDA MagikEDA is a python library which provides a set of functions to create graphs meant to be useful in Exploratory Data Analysis (EDA). It is conceived to be at one layer above libraries such as Pandas and Basemap. For now it is nothing more than a crude collection of some functions I needed to create in the last few months and decided to move to a public repository. Luckly there will be here something somone will find useful. Any doubt, do not exitate in contacting me.
There are currently three packages available. The packages univar and bivar have functions for plotting one and two variables, respectively. The package geo has functions for calculating distances and plotting maps. The functions included are able of executing the following:
- Plotting a “data frame profile”, which consists in a figure having one subplot for each variable in the dataset. Each subplot shows the variable distribution (histograms for numeric variables and bar charts for categorical).
- Plotting scatter plots with a tip box that is shown when points are hovered.
- Function to add an additional x axis bellow a figure.
- Function to compute the distance between a point and a segment (composed by two points).
- Function to plot a path (composed by a list o coordinates) over a map (the function automattically chooses a viewpoint to show the path (optionally with a padding)).
For documentation, see: http://magikeda.readthedocs.org/en/latest/