You can pass a list of values in the URL to "feed" a filter. But in the example you gave, you would have to generate 195 items in that list.
There is a nifty example out there of building such a list:
But maybe that's not the way to go for your example. Maybe it would be easier to re-generate that list in the target dashboard. Just pass the histogram bucket value, and in the destination workbook use both data sources, generate the value and blend on the added data source.
I'm just thinking out loud here... maybe keep the current data source in the target sheet as the primary. But in the secondary re-generate the histogram buckets, but in a calc(in the secondary source) only set some on-off value for the target bucket. ("On" for the target bucket, "Off" for the others). Blend on Site ID, and only keep sites in the primary where the calc is ON in the secondary.
It's just the seed of an idea.
Thanks for the response.
Blending the data sources is not possible without significant performance lag. I had to use a dummy set of data for this post, but the real data sets contain information for every radio of every Verizon cell tower in North Carolina, Tennessee, and South Carolina. The data integrity isn't great either, which causes additional performance lag when you do Tableau blending.
To get around the lag and allow users to quickly navigate between dashboards, it would be far faster to allow all of my workbooks to filter one another based on record ID. This is why I'm targeting URL parameter filtering.
I'm trying to give your previous value idea a shot, but it doesn't look like it's working the way I'm expecting it to. Each bar of the histogram represents a bucket of a Tableau Bin, and for each bin, there is a COUNTD(RECORD_ID) calculation happening. What I need is for the output field to capture all of the data going int each aggregate calculation, but list agg the record ID's instead of counting them so it is action filter friendly. The result should be different for each bar: The left most bar has a list the list of 195 ID's it contains, the next bar has the list of its 293 IDs, etc.
I notice in your example that the result is the same all of the bars. Using your setup, what I'm talking about would show these for each bar:
Furniture: Small Box, Jumbo Drum, Jumbo Box, Large Box, Medium Box, Small Pack, Wrap Bag
Office Supplies: Small Box, Jumbo Drum, Large Box, Medium Box, Small Pack, Wrap Bag
Technology: Small Box, Jumbo Drum, Jumbo Box, Large Box, Medium Box, Small Pack, Wrap Bag
Does what I'm saying make sense?
I had this conversation with someone else some time back, and I still have the modified workbook from that.
The example I pointed to is just concerned about getting all the items on the sheet for the purpose of displaying it in the title.
See attached for a different treatment.
I am telling the table calc to restart with each department in this one.
Thanks again for all of the help, Joe.
I experimented this morning, and unsurprisingly, Tableau won't pass the huge lists into the URL parameter filter. I still love the trick you shared, though.
I also like your idea about re-generating the buckets, but both data sources are aggregated to different levels. The data source for the histogram has daily performance data, but workbook I'm linking to has 14 day averages inside. Because the time dimension is not present in the other workbook's data source, added the histogram bins to that source would multiply the amount of records.
Working on this has made me realize something: The problem I'm working on shouldn't exist. My current system is built with multiple data sources and Tableau workbooks for the same data set. By doing this, my team and I are enforcing multiple analysis methodologies for the same data set, which isn't effective. I'm going to go back to my team to see if we can consolidate the methodologies into one that works off of one normal data set. We need to stop re-inventing the wheel.
1 of 1 people found this helpful
Sometimes the root of viz problems starts with the shape of the data, so I agree with trying to re-form the data to enable the goal you are after.
All the same, the other stuff we discussed are worth hanging onto. Keep it in your bag of tricks. It will serve you well some day.