Challenge #24: ARIMA Time Series

[OldDogNewTricks]

I have attached my solution, I ended up with more accurate results than the original solution.  I'm guessing because I'm using the newer version of Alteryx which comes with an updated version of R and I'm assuming updates to underlying default assumptions.

When I complete all of the challenges I'll come back to this one and do it in the R Prophet package and see how the results differ.

[danilang]

I love it when a plan comes together

Spoiler

Dan

[Kenda]

Spoiler

My approach seems pretty similar to others who have posted.

[kat]

Spoiler

Numbers are a tiny bit off...

[MarMu]

Great forecasting exercise!

I'm not able to replicate the exact same result and the ARIMA-tool returns an error as one can see on the snapshot in the spoiler-section. I suspect that the ARIMA and forecast implementation has been changed in a later version. However I managed to hit the forecast correctly down to the 11th decimal. I conclude that it is close enough. The confidence intervals, however fluctuates +-1 unit(s). I tried to download the official solution and that has the same issue.

Spoiler

[OllieClarke]

I ended up with slightly different results, but I think that's due to using a different version of Alteryx, as I got the same results as the solution workflow output.

Spoiler

Spoiler

[jevon]

Slightly different values to the output file given in the start file. Downloaded and ran the solution workflow and came out with the same values as I did. Different values most likely due to a difference in the versions of the Arima tool being used.

Spoiler

[Inactive User]

I couldn't get the numbers to match exactly. I opened and ran the solution on my computer to compare to mine and it didn't output the exact same as the given output. So I think this is close enough :) 

[StevenBetts]

Here's my solution! :)

Spoiler