Using ProfileStage is it possible to perform column level analysis against Multiformat Flat File of variable block and having atleast 5 different layouts.
Any help on this front will be highly appreciated.
Thanks,
Profile Usage against Multiformat Flat File
Profile Usage against Multiformat Flat File
Data Mover
Hi,
Well Basicly No (not yet at least)
Profile Stage meta data comes from seuential file configuration file generated by sequential File ODBC connection (can't recall the file type)
Or from a DB connection.
It has no means to get a pure CFF meta data.
Hawk release offers a shared repository that will enable meta data sharing between all the tools (Profile/Quality/Meta/Data/... Stage)
Though even with that I'm not 100% sure it will work.
When you really think of it Is there any scence of profiling all the different formats together
I think not
To get real valueable understanding of your data, In My Humble Opinion (IMHO), you should seperate the different formats to seperate files and profile them seperately.
Think of it:
1 rows has 20 columns
1 row has 35 columns
1 row has 10 columns
and so on...
How would you show or would like to see percentage of null columns for the 15th column in the file?
Does it have any relevance in context with the data of other types of records?
IMHO, your better off keeping things simple then complex, so get the file split to 5 different sequential files with 1 fomat and profile each according to it's structure.
IHTH,
Well Basicly No (not yet at least)
Profile Stage meta data comes from seuential file configuration file generated by sequential File ODBC connection (can't recall the file type)
Or from a DB connection.
It has no means to get a pure CFF meta data.
Hawk release offers a shared repository that will enable meta data sharing between all the tools (Profile/Quality/Meta/Data/... Stage)
Though even with that I'm not 100% sure it will work.
When you really think of it Is there any scence of profiling all the different formats together
I think not
To get real valueable understanding of your data, In My Humble Opinion (IMHO), you should seperate the different formats to seperate files and profile them seperately.
Think of it:
1 rows has 20 columns
1 row has 35 columns
1 row has 10 columns
and so on...
How would you show or would like to see percentage of null columns for the 15th column in the file?
Does it have any relevance in context with the data of other types of records?
IMHO, your better off keeping things simple then complex, so get the file split to 5 different sequential files with 1 fomat and profile each according to it's structure.
IHTH,
Roy R.
Time is money but when you don't have money time is all you can afford.
Search before posting:)
Join the DataStagers team effort at:
http://www.worldcommunitygrid.org
Time is money but when you don't have money time is all you can afford.
Search before posting:)
Join the DataStagers team effort at:
http://www.worldcommunitygrid.org