How to use the same source without duplicating entire table but selecting a few columns only

[Anonymous]

Hello everyone, 

This probably was addressed before but I didn't know what to search for. 

I currently have this 29.4 MB CSV file that I download daily and it is increasing day by day. It has about 130 columns. 

Sometimes I need to use the same source (same CSV file) to make a different table with only 6 columns to do some pivot/unpivot. 

The way I do it is to duplicate table and remove the columns I don't need, but I don't believe that's the most efficient and optimized way. As you can see in the screenshot below I need to load the 29.4 MB again (which loads very hard, used an unpivot for 44 columns there) and another 58.8 MB (which surprisingly loads instantly, I just made a pivot table there). And I believe I will make more and more tables like these in the future. 

Can you please tell me the best practice for things like this? Or the way you are doing it? 

Thank you very much!

[kcantor]

@Anonymous 

Instead of duplicating the table, reference the table. 

[Anonymous]

Thank you @kcantor  Is that it? 

I looked it online and the difference between duplicate and reference is that the reference doesn't include the steps I did in the original table but seems to still work as a duplicate. 

Would that solve my loading of 2-3 (or future multiple) times the same source? Or just save some memory usage here and there. I'm asking because I don't know if it's worth trying to reconfigure everything from the start, or just use reference from now on. 

[Anonymous]

Hi @kcantor I tested with one of the tables.

Referencing actually includes all the steps I did in my original table (which means extra columns to delete). I did all the steps in power query that I needed for the new table (Remove columns, pivot table, etc.) 

And when I refreshed it loaded the 60 MB again, the same as before. but this time not instantly like it did before. took a few seconds more. 

When duplicating I get the original source without any steps that I did in the original table. But still loads 60 MB after I did all the steps in power query that I needed for the new table (Remove columns, pivot table, etc.) 

My question is if I can select only the columns I need from that source so I won't need to load the entire source again and again or create a new CSV that I need to update daily?