# multi dimensional data visualisation

I have a multi-dimensional data set ['Age','Location','Address','height', 'BMI',...] used in the binomial classification task. What would be a good approach to visualise my labelled data (or training data) in a 2D scatter plot. My aim is to visualise samples belonging to the class labels in 2D representation, which otherwise exist in the multi-dimensional space.

I new to this field, please excuse my incorrect use of terms/tags.

INFO: I am using python for this task

• There is to little information here for any answer! one dosnt simply "do a visualization" of a dataset, one does so as the response to some specific question about the dataset. So, what is your question? (to the data) – kjetil b halvorsen Jun 5 '15 at 14:53
• apologies, edited my answer. – Segmented Jun 5 '15 at 15:18
• don't do exploratory data visualisation in python. use an interactive tool such as mondrian (theusrus.de/Mondrian), tableau, excel power view... – seanv507 Jun 5 '15 at 15:27

Since you want to view data by labels I recommend in R the ggpairs function from the GGally package (examples). In the mapping option map the aesthetic color = labelvarname and that will color the populations by the classification label.
With limitations on the types of plots, you can achieve the similar result in Python with seaborn and pyplot function pairplot.
The caveat to this approach is that if your number of dimensions / predictors increases, understanding of ggpairs output will be limited to the screen size as it outputs a matrix of plots.