wierd scatter plot points loop when plotting two features for regression

2
votes

I am trying to implement linear regression on my weather data with maximum temperature and pressure as features. I plotted them individually against time(time vs temperature and time vs pressure) and the plots seems fine however when I plot temperature against pressure (see the images attached) the plot is a closed loop of scatter point which worries me as I can't have a regression line in such a plot. moreover, the Pearson correlation value for max-Temperature and Pressure is 0.732.

I was wondering if I should plot time vs temperature vs pressure 3d plot and apply regression on that.would that help solving my problem? thanks.

Temperature vs pressure:
temperature vs pressure

Temperature vs time:
temperature vs time

https://drive.google.com/open?id=1HS23qzK3B9zVodYQOtJWZ10NeLQ6JTox

above is the link to data I'm using. It is in pickle format.

1
pythonmatplotlibmachine-learningregressionscatter-plot
Would you please post a link to the data? - James Phillips
I have posted the link to data I am using please check it out. @JamesPhillips - Rhythm Saparia
Thank you for the data link. Please see my answer below. - James Phillips

1 Answers

1
votes

When I plot time, temperature and pressure as a 3D scatterplot, it does not appear as if the data lies on a surface such that as 3D surface equation can be fitted to the data. Here is my 3D scatterplot:

scatterplot