nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Table: Not repeating values best method?

akk2

Basically, I have a table with that information (km and cost are aggregated data):<pre class="_prettyXprint">Group | Brand |drivers | Model | mileage | km | cost A | Honda | 3 | Civic | 5 | 100 | 300 A | Honda | 3 | CRX | 6 | 180 | 350 A | Zonda | 1 | CRZ | 6 | 20 | 50 B | Honda | 1 | Civic | 6 | 60 | 50 B | Honda | 1 | CRX | 6 | 40 | 50What I'd like to display:Group | Brand |drivers| Model | mileage | km | cost A | Honda | 3 | Civic | 5 | 100 | 300 | CRX | 6 | 180 | 350 A | Zonda | 1 | CRZ | 6 | 20 | 50 B | Honda | 1 | Civic | 6 | 60 | 50 | CRX | 6 | 40 | 50So that I can have to totals:Total: Honda drivers: 4Total: Zonda drivers: 1Total: Honda Civic: 160 kmTotal: Honda CRX: 220 kmTotal: Zonda CRZ: 20 km</pre>So basically, not repeat the information for each row of groups which are duplicated (group/brand/drivers), because this data gets repeated but must not be added and displayed. I'd receive for instance:Group A, Brand Honda, drivers 3, Model Civic, mileage 5, km 4Group A, Brand Honda, drivers 3, Model Civic, mileage 5, km 5Group A, Brand Honda, drivers 3, Model Civic, mileage 5, km 12Group B, Brand Honda, drivers 2, Model Civic, mileage 5, km 12Group B, Brand Honda, drivers 2, Model Civic, mileage 5, km 1 Which would basicall mean 2 groups; A and B, Group A has 3 drivers and Group B has 2 drivers. Total 21KM for Civic from Group A, and 13KM for Civic from Group B. Anyway, I know that the model is strange and is making the report difficult... I saw that you can set "Suppress Duplicates = true" in the advanced properties of the table's columns and rows, but that doesn't seem to suppress every time! Just the 1st time maybe? I don't know I can't seem to work out the logic... Any suggestion? Thanks! Oh, I tried using a table in a table for the right portion of the table (model/mileage/km/cost), but then it becomes incredibly difficult to align rows by rows and even columns by columns for tables inside the rows... Here's a sample of my weird ''Suppress duplicates'' results (just with numbers for example):<pre class="_prettyXprint">1 1 1 1 1 1 1 2 <-- first 3 should have been empty1 1 2 1 <-- first 2 should have been empty2 2 2 2 1 2 <-- Worked!!2 2 3 1 <-- first 2 should have been empty2 2 3 2 <-- first 3 should have been empty</pre>Just no obvious logic... (Note: these are in footers for the aggregate results which might be the cause of the issue...) [edit] It's not caused by data being displayed being in a footer, the result is the same in the details section of the table... The suppress duplicates just does not seem to work correctly in Birt 3.7.0...

Find more posts tagged with

Comments

BRM

In the main table you have to set a group which is Group/Brand/drivers. Then in a sub table to the right you repeat the data from the main table except with filters for the grouping data. To do the totals you want you can use aggregations with filters if necessary. See attached.

akk2

Ok the issue seems to be caused by group headers and footers, once these are removed (deleted instead of hidden), the remove duplicate seems to work correctly. Not sure how/why group footers affect it but removing them seems to fix it. Now my issue is that it works too well. For instance, drivers, if Group A has 3 drivers, and Group B has 3 drivers, well, it's duplicate values but in this case I'd want to keep them because they're for different groups! I guess using an hidden group footer to insert an hidden different value might do the trick... Probably why the headers/footers prevented the suppress to work.

Hans_vd

Did you consider using the <a data-ipb='nomediaparse' href='http://enterprisesmartapps.wordpress.com/2011/12/13/birt-drop-group-header-property-and-table-border-lines/'>drop property</a>.It is designed exactly for what you are trying to do.

akk2

Thanks BRM just saw your message, the example is in version 3.2 which I can't open open because it's not supported... But yeah that's what I tried to do at first, the issue with sub-tables is that it gets so messy and it's so difficult to align the different rows and columns... The different sub-tables for instance won't align their columns with each other... I guess I could use 'fixed' style and set every column/row width/height... Ok I'll try that if I can't get the above to work! Hans_vd: I've never head of that, I'll look this up thanks!

BRM

<blockquote class="ipsBlockquote" data-author="Hans_vd" data-cid="121692" data-time="1383244171"><div>Did you consider using the <a data-ipb='nomediaparse' href='http://enterprisesmartapps.wordpress.com/2011/12/13/birt-drop-group-header-property-and-table-border-lines/'>drop property</a>.It is designed exactly for what you are trying to do.</div></blockquote>That is cool. Never knew about that.

Hans_vd

<blockquote class="ipsBlockquote" data-author="BRM" data-cid="121705" data-time="1383253292"><div>That is cool. Never knew about that.</div></blockquote> I had been working with BIRT for quite some time before I learned about this.You don't run into this feature by accident it seems...

akk2

Hans_vd: Definitely! One of those hidden features.. Especially since it's in a header cell, and has more to do with detail but anyhow... I think this will definitely work! For those who may want to use it: It's not that intuitive though, if you have like 5 groups like in my case, you'll have to use something like this:<pre class="_prettyXprint">GrpHeader: xGrpHeader: yGrpHeader: z subtotal1GrpHeader: xxGrpHeader: yy subtotal 2</pre>Basically, you have to have each group header data in its own group header to repeat, and you can have totals, but in the headers, which will get dropped down in details and footers... You can have two values in the same group, as per total above, but you couldn't have everything on one row.. Anyhow, hopefully that makes sens to potential readers

akk2

Hmmmm.... I just ran into a problem though, how can you make totals using these dropped values? It seems that eventhough the headers are dropped, any sum uses the repeated data columns! So to go back to the original example:<pre class="_prettyXprint">Group | Brand |drivers | Model | mileage | km | cost A | Honda | 3 | Civic | 5 | 100 | 300 A | Honda | 3 | CRX | 6 | 180 | 350 A | Zonda | 1 | CRZ | 6 | 20 | 50 B | Honda | 1 | Civic | 6 | 60 | 50 B | Honda | 1 | CRX | 6 | 40 | 50What I would like to display:Group | Brand |drivers| Model | mileage | km | cost A | Honda | 3 | Civic | 5 | 100 | 300 | CRX | 6 | 180 | 350 A | Zonda | 1 | CRZ | 6 | 20 | 50 B | Honda | 1 | Civic | 6 | 60 | 50 | CRX | 6 | 40 | 50</pre>If I'm trying to make a total for the above, I get for drivers 9 instead of 5, because it seems the aggregation functions use the values instead of the dropped headers, which is logical.... But I'd need to sum the displayed values as per the 2nd sample... So far, I can't seem to find a simple way to do it... I need a total of unique values based on a second column. At worst, I might do an imbricated table for the totals which i'd group only two columns and display only group total? But I'm sure there's a better way... I think with filter somehow, if it would be possible to specify for instance, for the group 'brand', only row 1/0... Or only once per group... hmm...

kclark

Can you provide some sample data? I used the suppress duplicates feature in 4.3.1 and then aggregated some values in the footer of my table and it seems to be working as expected. I'm not sure why this would only suppress on the first n rows of a table.

Hans_vd

Hi Akk2, You could create a (invisible) column with an aggregation in it on the driver's group level (or the Brand level if Drivers is not a separate group). You can use whatever function like FIRST or MAX and then do a SUM on that field.

akk2

For sample data, not really it's a scripted datasource so getting it to work would be quite a headache, the actual data has nothing to do with cars/drivers... It might be a version issue, I'm using Birt 3.7.0, but anyway any data should work. For my issue, we've went a new direction due to increased report complexity so it's not really an issue anymore. And if it works in the latest version... hans; Yeah that probably could have worked.. What I ended up doing for some of those more complex aggregations is use specific imbricated tables to crunch the data for each subtotals. There's many ways the data for the totals have to be 'crunched' so the simplest way to do it really seemed to be passing a few values to a new table, then have a whole new sets of groups/filters to play with. The only trouble with that, sadly, is that when the report is displayed in a browser via HTML, the columns sometimes don't seem to match up very well between the main table columns and subtable columns... Oh well that's another issue

Thanks!