ARIMA R Model – Tableau Visual
Time Series Analysis and Forecasting has come a long way in the last few years. Today, We do time series analysis to forecast customer demand, future sales, inventory levels and much more. Building accurate models has become much easier and its actually fun now. I recently reviewed an on-line class that was promoting a collaboration tool; I wanted to see if I could reproduce the results first in R and then migrate the plots to Tableau. I’m not a Tableau partner or re-seller. I like Tableau and the public tool is Free. I love Free. Free means limited; However, I was able to build a respectable result in Tableau of a R forecast Plot and I wanted to share this with you.
If you are in school or a new start up with a limited budget; You can start off low cost / no cost. You can build some great machine learning and visuals without breaking the bank. Also, Keep in mind that my method for this does not include dynamically linking R and Tableau. This is a pipe-line solution. Think of data moving from one station to another and as it moved, the data changes. That’s important because I believe data pipe-line is the future. Data will always be moving rather than just being stored in one place.
This is the Repository: https://github.com/kyanyoga/timeseriesOne. The arima_ts_bsm10.R file is documented. If you are absolutely brand new to R; I would suggest taking a Time Series R class or reviewing documentation on Holts-Winters and ARIMA : Auto Regressive Integrated Moving Average. If you are a moderate to skilled R user, This should be straight forward. We are loading an excel worksheet that includes seasonality sales data. We’ll convert the data to a time series, do some plots and testing to understand the ARIMA model. We’ll plot the forecast in R and export the source data and forecast to Excel. Finally, Link the Excel data source and build a dashboard of the forecast in Tableau. There are several discrete steps but can be automated with collaboration tools like I mentioned earlier. Send me an email if you would like more information.
Check back next month. I’m building a magic square of data science platforms that I have evaluated that include that collaboration tool I mentioned earlier.
Here is the link to the Tableau Public Dashboard. Again, I could have spent hours on this Visual and included lots of fun elements but I wanted to this to be something easily duplicated and evolved quickly. Download the Workbook and Tableau public if you don’t already have a copy. Note: This is meant to be an example of what you can do with limited resources.
Have fun! Wow your friends – Predict your house-hold spend over the next few months and explain to your significant other that you can afford that new boat or sports car and pay the dentist or not :).
Email: firstname.lastname@example.org – Phone: 765.325.8373 ( call / text ).
Stay Informed: Please Contact US | Subscribe to our Blog Updates or If you would like more information.