Another hallmark of excellence! I continue to be amazed by the quality of these videos -- it would be hard to improve on them. Thank you very much for your phenomenal work!!
This video is amazing, I was reading the data warehouse toolkit and I love it, but this part was so confusing IMO. So I browsed around and found this, nothing left to clarify!
I really liked the way you explained the tables with real time examples.. I was looking for multiple videos to get the correct understanding but this has enough information... Thanks much
Very Nice ... Shot n crisp info... Have one question if there is a periodic snapshot fact table or accumulating snapshot fact table then there should be be a transactional fact table by default, ... Right?
Great video and playlist, thanks! Just an observation but when you are discussing Update frequency isn't it: Transaction - None Periodic - at defined intervals Accumulating - at each milestone
2 suggestions: 1) please order your videos in the correct way within the playlist 2) please go slow while explaining I have watched all your videos and liked the content.
Great question. Fact table type will be dependent on two factors- first is volume of data that we can expect. Most common criteria or key to store data would be unique visitor or visitor # or ID. Second factor would be type of analytics we want to do. Option would be periodic snapshot but I would go ahead and gather every detail just as transaction table. Again, thanks for your wonderful question.
Thx @aroundBI for your response, I didn't get your sentences here "Most common criteria or key to store data would be a unique visitor or visitor # or ID". As far as I understood the lowest level of granularity in case of the snapshot is the "date", in my case is 'sec'. so we will get a row per each visitor Id + time. this would be unique over the file. Then this looks like a snapshot, right? Are u confirming me? Mostly you are selecting snapshot policy just for avoiding of data grow in fact, right?
Can a periodic snapshot contains lowest level of granularity...In other words business wants to take a snapshot of the entire fact table on daily basis, so that on a particular day they want see all the relevant detail rows anytime possible..How to handle this.
If a Transaction has more than 1 product that have been bought/sold then in Transaction Fact Table Transaction number will be the Degenerate Dimension and same transaction number will appear twice as two different product FK are there, with different price and quantity for each product in Transaction Fact table. So, it has one row /transaction does not necessarily hold true.
How can a periodic snapshot fact table be less in size when compared to transaction fact? I have worked with snapshot facts which are in hundreds of millions whereas transactional fact are just in tens of millions.. can't say that is always less, as it depends on the grain
Another hallmark of excellence! I continue to be amazed by the quality of these videos -- it would be hard to improve on them. Thank you very much for your phenomenal work!!
Your feedback is big encouragement to continue our effort. Thanks a lot.
You're very welcome! Thank you for the kind reply! :o)
you are speechless, Very very good Explained, thank you
This video is amazing, I was reading the data warehouse toolkit and I love it, but this part was so confusing IMO. So I browsed around and found this, nothing left to clarify!
Same!
@@davida6146 Same!
I really liked the way you explained the tables with real time examples.. I was looking for multiple videos to get the correct understanding but this has enough information... Thanks much
Thanks, that explained a lot.
thanks for this meaning full explanation...........................................................................
Best Information, thank you Team.
Very Nice ... Shot n crisp info... Have one question if there is a periodic snapshot fact table or accumulating snapshot fact table then there should be be a transactional fact table by default, ... Right?
In my experience, I have seen a transaction table also along with periodic snapshot. Not a rule but common practice.
it will be great if u also add an addition where,, u can show how to add a surrogate key from customer/ product dimension as well.
very much useful sir. Thanks a lot.
Really good video series on datawarehousing concept..what type of fact table will you recommend for a CRM Saas application
Excellent
Thank you.
Nice video man, thx a lot.
Awsome !!!
Nice 👌.. I have a question- what is additive, semi additive and non additive?
Good explanation! Thanks!
thanks you again for these videos, u are helping many people.
Great video and playlist, thanks! Just an observation but when you are discussing Update frequency isn't it:
Transaction - None
Periodic - at defined intervals
Accumulating - at each milestone
Thanks for the video
Very well Explained, thank you
thank you and impressive, clear explanation
Thank you great one
2 suggestions:
1) please order your videos in the correct way within the playlist
2) please go slow while explaining
I have watched all your videos and liked the content.
you are speachless
Nice explanation, pretty clear, thanks
Thanks for your support.
No entiendo como lo pudiste entender, encima los subtítulos están desactivados.
very well Explained!! :)
Thank you ! Well explained!
another good video
really good tutorial!
Thank you for the video, I need a help, User tracking, such as Google analytics data, we need to keep track of the user, what type of fact table is?
Great question. Fact table type will be dependent on two factors- first is volume of data that we can expect. Most common criteria or key to store data would be unique visitor or visitor # or ID. Second factor would be type of analytics we want to do. Option would be periodic snapshot but I would go ahead and gather every detail just as transaction table.
Again, thanks for your wonderful question.
Thx @aroundBI for your response, I didn't get your sentences here "Most common criteria or key to store data would be a unique visitor or visitor # or ID". As far as I understood the lowest level of granularity in case of the snapshot is the "date", in my case is 'sec'. so we will get a row per each visitor Id + time. this would be unique over the file. Then this looks like a snapshot, right? Are u confirming me? Mostly you are selecting snapshot policy just for avoiding of data grow in fact, right?
@@maryam4071 🙏
well explained ... thank you
Can a periodic snapshot contains lowest level of granularity...In other words business wants to take a snapshot of the entire fact table on daily basis, so that on a particular day they want see all the relevant detail rows anytime possible..How to handle this.
is unit price a additive fact or non-additive?
Please add a video on late arriving dimension
Why you are not uploading video more than one year in you channel..?
well explained, thanks
Thanks. Our aim is to keep things simple.
If a Transaction has more than 1 product that have been bought/sold then in Transaction Fact Table Transaction number will be the Degenerate Dimension and same transaction number will appear twice as two different product FK are there, with different price and quantity for each product in Transaction Fact table. So, it has one row /transaction does not necessarily hold true.
Explained all in 9 minutes
why is the accumulating fact table not considered as a dimension table? it has the date dimensions.. same as role playing dimensions
What is historical data in data warehousing?
Hi need some help related dimensional modelling its urgent
what is difference between fact and fact table
How can a periodic snapshot fact table be less in size when compared to transaction fact? I have worked with snapshot facts which are in hundreds of millions whereas transactional fact are just in tens of millions.. can't say that is always less, as it depends on the grain
True
Good explanation, but you need to speak a little slower please
why dont you set play back speed of video to 0.75x...