Hello!
I have a flow that is getting complex, originally there were only about 40 columns, after being treated with several Joins, they reached the result of 400 repeated columns ("Left_something","Right_Right_something_else"), several of these columns are completely null or even with cross values (Half is on Left_ and the other half is on Right_).
I would like to merge them, I made a flow that uses a CrossTab, I used a regex to remove all "Left_" and "Right_" from the columns so that when it does a Transpose, it merges the columns of the same name (and as the null values were removed, I already got the first value anyway). It worked, but at the expense of the metadata of all doubles that turned into commas (eg 1,234.56 in the Brazilian standard became 1,234.56), in the final result, there was a mixture of decimals separated by dots and separated by commas in all columns with doubles.
I thought about converting the dots to commas, but they are mixed up in the same column, and have different decimal sizes. The metadata contains the true value.
How can I remove all spare "Right_" and "Left_" columns without breaking the Doubles?
An example of input data:
Expected result:
Is this the result of CrossTab:
But I get the result in a real flow (Data with commas and dots mixed up):
I put below an example of what I tried to do, but in the example it does not occur to truncate the doubles from half to a comma and half to a point, but in the flow that I am applying with thousands of data this is happening as in the last image above.
