Performance when importing tables with MANY columns

[vehau1]

We are working with files in csv-format, which are large and have very many columns (several thousand).

I have found that Power BI struggles a lot with this amount of columns, lags and closes, and thus, Spotfire is more suitable. Power BI performs much better with fewer columns and longer rows doesn't it? 

I usually reorganize the files by unpivoting, so that I get the data in a large "stack" and reduce the amount of columns greatly. While this way of handling the data is fine and can provide many possibilities, I miss some possibilities which I had with the original setup. Is anyone familiar with this issue, and is it in any way possible to improve the Power BI performance/user experience when handling large-number of columns files? Also, what is the capacity limit in different visuals? I loaded a bunch of values into a normal line chat and put a legend on them, but it couldn't show me all the data at the same time. I have Power BI Pro.

[v-xuding-msft]

Hi @vehau1 ,

When we connect with import mode, it does have some limitations so that the report will load slowly. You could reference the blogs to try to optimize your data model and improve the Power BI Performance.

Power BI performance best practices 

A comprehensive guide to Power BI performance tuning 

POWER BI PERFORMANCE: SIMPLE PRACTICES FOR REPORT OPTIMIZATION 

introducing the power bi performance analyzer 

For visuals limitations, please reference the document : Data point limits and strategies by visual type 

For your line chart without showing all of data,  I think it might be caused by the context of your data. If it could drill down, you could expend the rows or columns to show completely. If it couldn't, please try to remove some data to check if the visual works fine. And then create some measures or calculated columns to get the results that you want.

Best Regards,

Xue Ding

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly. Kudos are nice too.