Project research problem - Help with model and Alteryx method needed

[DMI10000]

Hello all,

My name is Dave, and I am trying to solve a problem but not sure how to approach it or the method to use in Alteryx.
I have weather data (temperature, pressure, humidity etc) and Google Analytics data and I am trying to investigate or find a correlation as to how the rise in temperature may impact the uptake for air-conditioning units.
1. What model/method do you suggest I use in solving this problem?

2. What tool in Alteryx can be used to approach or solve this problem?

3. Do I need more data sources?

4. Any other valuable responses will be welcome.

Thank you.

[grazitti_sapna]

Hi Dave,

There are 2 methods in Alteryx which we can use for finding the correlation between 2 variables:

1. Pearson: You can use this method if the magnitude of difference between the variables is small. For Example: Age and Years at Job
2. Spearman: If the magnitude of difference between the variables is high, then you can use spearman. For Example, the values of salary are usually higher  in magnitude than years at job.

Another solution which I would recommend is using Heat Map tool to show correlation between them.

[DMI10000 · Accepted solution]

Thank you for the reply and info and it was very helpful. 

The other question I want to ask is:\

1. How do I measure or with what tool do I use to identify or understand at what temperature threshold do I begin to see a significant difference in customer uptake due to the temperature increase.

Thank you.

David

[grazitti_sapna]

@david In that case you can use an interactive chart tool, in this tool you will get the values plotted in a line form and then you can check the point where its showing a significant difference in customer uptake corresponding to the increase in temperature.

[DMI10000]

Hi Sapna,

Thank you very much for the suggestion. However, I think using a visual chart as you suggested will be best as a validation step. Also, a charting tool will not give a central point of convergence for deriving the threshold temperature that influences customer uptake. My thought is that we need a statistical means of establishing at what threshold temperature an uptake is significant.  

What do you all think?

Dave.