User Input Breaking my workflow

[jipillo]

I have a workflow that functions properly (the one attached) until I change it into an app so that i can have a more user friendly process for changing data.

The Input on the far left accepts a File browse and Action update with no issue. The two on right, named Test and Cluster break once the input is changed. 

Through the process of elimination, it looks like once the input is changed the joins dont recognize the formats for the joins. Looking through the output log, it will read the input data from the new user input, but something is happening so that will not join the data.

Also, once its broken there is no going back. I have to revert to an earlier saved version and start over.

The end goal is turn this into an app so that the end user doesnt have to know anything about Alteryx to run it on their own computer.

[AbhilashR]

Hi @jipillo, a few questions to help debug the issue:

• The workflow you provided only uses Input.xlsx in all of its Input tools, while you shared 3 separate excel files. Is that how it needs to be? If yes, then the issue could be with the Test and Cluster tabs of Input.xlsx file. They are empty.
• If not, then try using the three different files in the three different inputs and see if that works.
• Also, can you confirm if the datatype for STORE_NBR column going into the Join tools are the same? It is a Double for the excel leg of the input going into the Join tool, what is the datatype coming from the Dynamic Input tool?

[jipillo]

You hit it right off the bat.

When I manually changed input to the future one (the ones i attached), it broke and was gving me Str_Nbr Missing as an error. I added a Select to the Dynamic input to see what Store_nbr defaults as an output. It is a fixed decimal. Should i change them both to double? Or would that have no effect.

Edit - Also it is inconsistent, some times its fixed decimal and sometimes its int32. Must be the different tables.

[AbhilashR]

I am inclined to say it shouldn't matter in this case since store number values are integer in your case, but do profile the join outputs to ensure you aren't losing any records.

[jipillo]

I am testing the change to double with the new inputs and see if it breaks it.

The one that doesnt have a problem joins on a name that is a string. If what i am currently testing doesnt work, I am going to try converting the number to a string.

[jipillo]

Changing to a double had no effect. I am going to try String.

[AbhilashR]

Hi @jipillo, were you able to resolve the join issue? just curious. 

[jipillo · Accepted solution]

i finally got it to work. I had to format the number to get rid of zeros and then turn it into a string before it would join them. 

Long story short, joining on a double caused the problem.

[AbhilashR]

I was thinking about excel dropping leading zeros for numerical values which could have caused a potential join issue. Glad it worked out and thank you for the feedback.