1 2 3 Previous Next 31 Replies Latest reply on Dec 10, 2019 1:12 AM by Bangaru Saivivek Go to original post
      • 15. Re: Data Extraction
        Bangaru Saivivek

        Hi steve,
        1)All tables are of different size

        2)may be 100's of tables

        3)New tables are added

        • 16. Re: Data Extraction
          Bangaru Saivivek

          How should i contact you

          • 17. Re: Data Extraction
            Steve Wood

            Hi Bangaru,

             

            Thanks for the extra info - that clears it up in my mind quite a bit; particularly if clarification 1 = the tables have quite different columns.

             

            I'll be interested to hear whether others have ideas for you, but based on my limited knowledge the JavaScript API isn't what you're after. I had a a quick look at the Extensions API too (which looks really cool). Reason I don't think these are what you're after is that I didn't see API methods for changing data sources and adjusting the dimensions/measures in a sheet. Still, others may comment more positively. I'm really no expert there!

             

            If you were going to build a JavaScript based solution - presumably hosted on one of your web servers - I'd be really interest to hear why you want to involve Tableau in the solution. The reason I say that is that if I've understood correctly you want to let users pick a database table name and then show the data in that table. You could probably just write a web app to do that without involving Tableau. What Tableau functionality where you hoping to include and have your users benefit from?

             

            If the answer to the above question is filtering, then my mind comes back to an ad-hoc query tool. Or even providing your users with something like SQL Server Management Studio and fairly restricted access.

             

            If the answer is that you want your users to be able to do various analytical tasks like SUM measures up, slice and dice by dimensions, or visually explore the data, then I suppose we should ask "why not just give them a Tableau Desktop licence and access to the DB tables?". What's the reason not to go that route out of interest? Licencing costs?

             

            If you want to get really hacky then could you potentially write a stored procedure / view / query on the database in question to pull data from the required tables and restructure it into a consistent output - e.g.

             

            SELECT 'tbl_blah' AS tbl_name, 'header' as row_type, 'id' as pk_or_id, 'some_value' AS col1, ...

            UNION

            SELECT 'tbl_blah' AS tbl_name, 'data' as row_type, CAST(id AS VARCHAR(MAX)) as pk_or_id, CAST(some_value AS VARCHAR(MAX)) as col1, ...

            FROM tbl_blah

            UNION

            SELECT 'tbl_user' AS tbl_name, 'header' as row_type, 'user_name' as pk_or_id, 'user_height' AS col1, ...

            UNION

            SELECT 'tbl_user' AS tbl_name, 'data' as row_type, CAST(user_name AS VARCHAR(MAX)) as pk_or_id, CAST(user_height AS VARCHAR(MAX)) as col1, ...

            FROM tbl_user

            ...

             

            I wonder if you could automate the generation of such a proc / view / query from the DB catalog so that new tables of a particular pattern get added on a nightly basis? Once you'd got data in a format like the above then some of the earlier suggestions become a possibility - i.e. just use natural Tableau filtering. It is pretty hacky though so I don't know whether I'd go there! You'd need a way to show a fake header row (row_type = 'header' above); and you'd probably want a way to hide columns that weren't applicable to some tables...

             

            Well that's all I've got right now. Good luck!

             

            Ta,
            Steve.

            • 18. Re: Data Extraction
              Bangaru Saivivek

              I am sharing u one excel sheet all info i have mentioned their wait a minute i will attach

              • 20. Re: Data Extraction
                Steve Wood

                Hi Bangaru,

                 

                Okay, that looks like it could be quite different to the picture I drew ... advice may change!

                 

                When I look at your screenshots I think I'm seeing groupings of metrics rather than tables, and the ability to filter to an organisation group and date range for which we want to see those metrics.

                 

                I'm interpreting GROUP (e.g. QUALITY) and REPORT (e.g. AFC BILLING) as groupings / categories of metrics. Within the AFC BILLING category I think I then see 3 further sub-categories: AFC BILLING; AFC BILLING DETAILED; and AFC BILLING PERCENTAGE. Within each of these sub-groupings I then see a list of metrics. E.g. Near Misses, or % OF CASES PENDING blah.

                 

                It isn't clear what should be shown in the space where you've made some comments. I get the impression that someone will have selected a group and a category, and will then see the list of sub-categories and metrics. They then click on a specific metric and see ... what? E.g. what do they see when they click "Near Misses"? You mentioned rows and columns above, but it feels more like they'd either see a single number - the number of near misses in the period selected for the organisation group selected - or perhaps how that number has changed over time, e.g. a line chart of near misses by month. You tell us though, rather than me make more incorrect assumptions!

                 

                How is the source date structured? And can you shape it (e.g via some ETL) to make some of the Tableau work easier?

                 

                 

                Ta,

                Steve.

                • 21. Re: Data Extraction
                  Bangaru Saivivek

                  when we click on Metric value(i.e. Near Misses) behind it has sql query from oracle DB we have to run that query and show the results of different rows and colums in that space

                  • 22. Re: Data Extraction
                    Steve Wood

                    Hi Bangaru,

                     

                    Final thoughts from me below, but first up: If those screenshots aren't obfuscated / example data then you may want to remove or update them to hide names.

                     

                    This is a really interesting requirement and perhaps one that doesn't really suit Tableau, as I guess you know. I'm curious as to what your end users will do with the tool? It's not really a dashboard for decision making or analysis from what you've described and posted. E.g. I'd have to do quite a bit of heavy lifting as an end user to work out if near misses were increasing, and what if any patterns seemed to be driving that (e.g. specific locations). Given the Excel download option in the screenshot I wonder if people are using the tool to download data to analyse it in Excel. Do you think that's the case? Or are there other usage scenarios where the list of records helps?

                     

                    If the Excel downloads are a clue that the analysis is happening outside of the tool, then you could dig into what analysis is being done in Excel; what graphs do the end users produce regularly; what export options do they not even use? Given the breadth of data included it could easily be the case that lots of it isn't used right! Once you have that info you might have some good requirements for a cut down set of dashboards where Tableau would add a whole heap of value. I realise this could be a lot of work, but felt that I should mention it.

                     

                    Ta,

                    Steve.

                    • 23. Re: Data Extraction
                      Bangaru Saivivek

                      This type of dashboard already done using java

                      we need it to be done in Tableau along with other Tableau projects

                      • 24. Re: Data Extraction
                        Steve Wood

                        Hi Bangaru,

                        Maybe someone else will post with something I'm missing. Otherwise you could try the Tableau support or feature request suggested by one poster above. I genuinely think you'd make headway by digging into what the end users do (or plan to do) with such a dashboard - if you haven't already done so - but do appreciate that that could be a big job, and potentially outside of your scope. Alternatively you could try finding one of the key users of the dashboard and giving them a trial Tableau Desktop licence (plus mentoring) to see if any light bulbs switch on for them that there is a better way with Tableau. If i think of anything else I'll post back, otherwise best of luck with this one!

                        Ta, Steve.

                        • 25. Re: Data Extraction
                          Bangaru Saivivek

                          Ok Steve Thank you, If you found anything related to this query please give me some help.

                          • 26. Re: Data Extraction
                            Bangaru Saivivek

                            Hi steve 1 more help,  'Tableau Book' is the default font i have to change the default font in preferences how to do that?? 

                            • 27. Re: Data Extraction
                              Steve Wood

                              Hi Bangaru,

                              It's probably best to post that as a separate question. That way when you mark comments as helpful or correct, and people search in the future, the info will match up to the question. It can get quite hard to track things when asking unrelated questions on an existing thread. Best of luck.

                              Steve.

                              1 of 1 people found this helpful
                              • 28. Re: Data Extraction
                                Bangaru Saivivek

                                I have mentioned in new topic no reply from you