R: Fitting a Model

Learn how to fit a model in R, plot the smooth curve over the data, using lm() and predict() functions.

[@loki92loki]

Thanks, Help me a lot

[@g.greyzone7300]

This is so good and exactly what I was looking for. Thank you so much. Why, oh why, does the person who does something so good, stop here. Seriously sad :(

[@darthszarych5588]

This was helpful thank you!

[@waleros]

Nice! Thanks! Although, do you know any methods on how to determine the proper equation based on the shape of the plotted data?

[@manyminimoose1118]

My pleasure! The proper equation can be difficult to determine, and in general it comes down to guesswork and experience. If you know more about the process that generates your data, then you can use that to get a closer approximation. Something you can play around with are more complicated formulas, such as y~log(x), and composite formulas such as y~exp(x)+x^2+x.

[@manyminimoose1118]

If you want a method that doesn't require you to guess the proper equation, you might try something like "loess" (stat.ethz.ch/R-manual/R-devel/library/stats/html/loess.html), you can use it in a similar way to "lm", and it will work with "predict". Try running the example code at the bottom of the link.

[@waleros]

Awesome! Thanks a lot!!! Keep​ it going!!!

[@manyminimoose1118]

Any time! I'm moving my educational content to a new channel (ManyMiniMaths, goo.gl/4UMWZt) so if you want more similar content you should check it out! I'm also taking suggestions for topics you'd like covered (comment on a video or tweet me @mrparker9090).