Suppose, I want to create another analysis_2 using only T4 & T5. I about choose
New Analysis
but here it is showing Joined symbol T4 Dataset – pls see screenshot ( It contains T4, T5 & T6 DATASETS which are used in Analysis_1 ). But i want to take first T4 for the analysis and add T5 dataset in it for the joining purpose. ( I knew it I can create the analysis_2 using same joined dataset but i am not).
I can take duplicate of the Joined T4 dataset and I can remove T6 dataset from that duplicate joined dataset.
Why we need joined datasets rather than individual datasets. Can’ t I take simply individual datasets for the different Analysis and i will add extra datasets if i want?
T4 – is the Driving table/dataset or you can say Fact table
Analysis_1 ( T4, add T5 & T6 ) for the Dashboard_1
Analysis_2 ( T4, T5, T7 & T8 ) for the Dashboard_2
Analysis_3 ( T4, T6 & T8 ) for the Dashboard_3
Because i don’t want extra datasets/ tables in my Analysis purposes. Moreover, I feel creating duplicated joined datasets or datasets is unnecessary.
Apologies if I hurt anyone with stupid questions
If I put a schedule refresh for the Dashboard_1, i need to refresh only or it will trigger only on T4, T5 & T6 tables for the latest data, right? Why i need to put extra effort on other tables to load from a DB or anything to the Dashboard_1
regarding you question joined vs individual: Yes you can use individual datasets and can add different datasets in a analysis. But if you want to have data of T4 and T5 in one visual, it will not work with two seperated dataset. therefore you would need to join T4 and T5 as one dataset.
regarding you question data refresh: if you only want to update dashboard_1 data, yes. only update T4, T5 and T6.
General: you could build one dataset per datasource as a master dataset everythingelse you do as a slave of masters.
e.g.
T4_m = Table4, T5_m = Table5, T6_m = Table6 → Dataset_4_5_6 (join T4_m, T5_m, T6_m)
T4_m = Table4, T8_m = Table8, T6_m = Table6 → Dataset_4_6_8 (join T4_m, T8_m, T6_m)
so every change in T4_m will go to the slaves
But I it would be good to have individual datasets. like which Analysis using which datasets simply so that in future everyone will care about datasets before deleting them with creating again & again like duplicates
Hi,
we can see the usage of dataset in analysis, by just clicking on usage in dataset. Also, you can user proper naming convention while creating datasets to avoid such scenarios.
