Python Tutorial : Visualizing Linear Relationships

Key Takeaways

before building models it is useful to explore your data visualization is an important part of exploring data as it can detect qualities of the data the summary statistics might miss visualization is also a compelling method of communicating your data and modeling results to others in this lesson we'll review the use of MATLAB to visualize data for some common examples of linear models when matplotlib was created it was motivated in part as an open-source alternative to MATLAB so the original interface was similar for example we have some data stored as numpy arrays x and y that we want to plot we start by importing matplotlib pi plot using the conventional alias flit second we pass the data into the function flit dot plot the minimal inputs are X comma Y but by adding a third input the string R - o you can set the plot style in this case r is for red the dash is for a solid line and O is for the round data point marker a complete list of options are provided by the matplotlib documentation lastly we use Pluto to display the plot created or else you'll see nothing there is another approach to using MATLAB which is more object-oriented at first it may seem like more effort but it is vastly more customizable and for complex plots easier to use again we start by importing pipeline from matplotlib but rather than calling the function plot plot we instead use plaits subplots to construct two new objects the figure object and the axis object notice that instead of calling it a function we'll use the term method when the function is part of an object or a class for ease of reuse we create a dictionary to store some of the style plot options we'll use to customize our plot then we call the access object method access stop plot to create a line object passing in both the data and the options dictionary using star star and packing to transform dictionary key value pairs into keyword arguments input to the plot method just as if you tuck them in then we use methods set Y label and set X label to add text labels to the axis object notice the Python convention of assigning any unused output to the underscore and finally we call it dot Show as before to display the constructed figure notice that there is no fig show method once you plot your data you may see a linear relationship how can you connect the plot to the ranges of values recall how we computed the ranges and the speed in a previous exercise start at the point x1 comma y1 equals 0 comma 0 and rise up 3 and run right to to arriving at point x2 comma y2 equals 2 comma 3 and the plot shown as the X variable increases to the right the Y variable goes up the change in Y the Green Line is dy equals y2 minus y1 equals 3 minus 0 equals 3 the change in X the red line is DX equals x2 minus x1 which equals 2 minus 0 equals 2 the slope of the line is seen as the ratio of the increase in Y divided by the increase in X or 3 over 2 the y-intercept of the line is the y-value where x equals 0 in the following exercises you'll get more practice using MATLAB to visualize your data and a model and even use plots to make rough estimates of model parameter