My Home Energy Estimator

My Home Energy Estimator is a personal tool which you can use to estimate future home energy usage based on your electricity bill history.

Try it at this link: https://zaros.shinyapps.io/my-home-energy

I used my home electricity bill history and annual local weather data as a starting point. The data points are few, only 10 observations since I had just moved to my location. The weather data was sourced from a non-authoritative weather site.

#The data looks like so:
knitr::kable(head(useData,5))

Using my learnings from this course, I can now explore the ways I can relate localised weather and personal activity data to my electricity bill.

For example, one of the models I am experimenting with is a Poisson Regression model which predicts on some weather data and also how many days I am at home.

glm(formula = Energy.Usage ~ Avg..Temp. + Precipitation + Air.Quality + 
    Days.of.Occupancy - 1, family = "poisson", data = useData[-(1:3), 
    ])

                       Estimate   Std. Error    z value      Pr(>|z|)
Avg..Temp.         0.2098223703 0.0062331937 33.6620967 2.074422e-248
Precipitation      0.0019614710 0.0004487509  4.3709575  1.237029e-05
Air.Quality        0.0003269131 0.0001886486  1.7329208  8.310975e-02
Days.of.Occupancy -0.0023125140 0.0056290391 -0.4108186  6.812055e-01

It would be fun and beneficial to continue this project. The electricity bills are inevitable. Now that I can apply data science, I hope to improve my tool as I collect more data and work towards scientifically managing this expense.

Perhaps I can improve data collection and prediction capability by installing a personal weather station at home and survey my neighbours for some comparative data.

See my project at zaros.shinyapps.io/my-home-energy.

Cheers!

zaros @hackplay.net