merge with InnerJoin

[CyrilMoysset]

Hey, I want merge (with Table.NestedJoin(...JoinKind.Inner)) a sample of 300k people with a page of 1 million people for example but time to merge is too much long... I have some questions about merging columns :

- For calculate is it better to choose numerical format (123) than text format (ABC) or it is not relevant ? 

- DAX language in Pivot is more effective than M language for bigadata ? if so what is the efficient calculation ? 

- is it a RAM problem of computer ? is better using server ? SQL server ? 

- Exist a race version of PowerBI on a server to avoid PowerBI desktop's problems ?

- Can I ajust settings to earn calculation speed ? 

[MattAllington]

Where are your tables coming from?  Are they both from a DB like SQL Server?  If so, you should make sure you only used the power query UI to create the joins, and check for query folding. Read about that here http://radacad.com/not-folding-the-black-hole-of-power-query-performance

numeric or text hsould make no difference. 

DAX and the power query formula languages do different things. You can't really compare them. It is best to shape your columns in power query if you can

a suitable server may work faster than your PC, but if the query is poorly structured, it may not help. You need to understand the problem regardless

there are no magic settings - you need to focus on query folding. 

[CyrilMoysset]

Thank you for your response 

The article seems me very interesting I'm going read this 

I want transform facebook data for marketing issues, I can use Query for that or I can use my own system linked to facebook API where data extracted is in database MySQL (phpmyadmin) but i don't known is one or the other is more adapted to transform data ?

In any case a paying version can help me ? 

[v-qiuyu-msft · Accepted solution]

Hi @CyrilMoysset,

If I understand correctly, it seems that you want to retrieve data from Facebook, right?  There are some samples you can refer to:

Tutorial: Facebook analytics using Power BI Desktop

Facebook page insights with Power BI

And about your original questions, I agree with @MattAllington, and would like to deliver my opinions:

- For calculate is it better to choose numerical format (123) than text format (ABC) or it is not relevant ? 

It depends on which results you need to get, as there are text functions and math function in DAX and M query for us to calculate. It would be better to choose the data type in a specific scenario. Sometimes, those two data types can be converted, see: Data types in Power BI Desktop.

- DAX language in Pivot is more effective than M language for bigadata ? if so what is the efficient calculation ? 

I would suggest you to know the What's the difference between DAX and Power Query (or M)?. As DAX uses functions to work on data that is stored in tables, while Power Query (M) is querying data sources, clean and load data. It would be better to shape data in Power Query.

Best Regards,
Qiuyun Yu

Best Regards,
Qiuyun Yu

[CyrilMoysset]

Hey thank you for your response 

I already read these tutorials and I know the method for little tranformations. I have marketing issue in my job and I would work with social networks big data, nowadays I have a merging problem but tomorrow I will have difficulties with visuals. I would like use the network graph but it loads very few data

for my main problem, merging : I will turn to SQL solution 

I note your informations, thank you