nickelcap Tracker

"(Blank)" in Card Visual For 0 Count

nickelcap — Fri, 26 Aug 2016 15:27:16 GMT

Hi Everyone,

I'm not sure if I encountered a bug or if I'm doing something incorrectly. I put together a card visual that provides a count of the assets that meet a certain set of criteria. I'm using three columns for the visual: "Asset", "Colour Formatting", and "Latest". "Asset" provides the asset name; "Colour Formatting" reports a 0, 1 or 2 based on certain criteria; and "Latest" finds the latest time stamp for each asset. If that's not clear, here's a snippet of the data:

And here's a snippet of the set up for the visual:

The count is set as Count (Distinct). Now, the filters I'm applying will mean that there will be 0 assets in the count (I know this by manually checking), but it shows '(Blank)' in the card. I know this set up works because I'm using something very similar for another count, the only difference being that Colour Formatting is set as 2. Is there a workaround for this?

Thanks,

Dan

Re: Table Column Total Incorrect

nickelcap — Wed, 27 Jul 2016 16:48:34 GMT

Vvelarde I had wondered about that, but even with the standard count the total still is off. Would the duplicate names still throw off the total with the standard count?

Re: Table Column Total Incorrect

nickelcap — Wed, 27 Jul 2016 13:56:08 GMT

Thanks for the offer, kcantor. Here's a snippet of what my data looks like, along with the formula used to calculate the last half day:

From that sample, in the table visual 'Name' is counted using the 'Count (Distinct)' option and BASE is set as 'Don't summarize'. I then bring the 'Last Half Day' column to 'Visual level filters' and filter to only 'Past Half Day'. Currently, I'm getting something that looks like this:

The total I'm looking for would be 27, not 26.

Thanks,

Dan

Re: Table Column Total Incorrect

nickelcap — Tue, 26 Jul 2016 16:51:33 GMT

Thanks for the link. I think I have something of an understanding of what's happening after, I'm just not sure that the solution offered in that post helps me solve the problem. I'll have to work at it.

Thanks,

Dan

Table Column Total Incorrect

nickelcap — Tue, 26 Jul 2016 14:39:30 GMT

Hello,

I've encountered an issue totalling a table column in Power BI. I have two columns, a category column and a count column. I want to have the sum of the count column below the table, but it comes up incorrectly. Here's a capture of my table at one time:

And here it is later:

As you can see, if the total were just a sum of the 'Count of Name' column it'd be returning 32 rather than 28 in the first example, and 29 instead of 28 in the second example. For a bit more information, the table has a filter applied to it such that only dates within the past half day are counted. The individual values under the 'Count of Name' column are correct. When I try just having a card for 'Count of Name' with the date/time filter applied, I get 28 as well. So there's something about the number 28 that Power BI is stuck on.

The way the time filter works is by creating a calculated column with the following equation:

Last Half Day = IF(AND('SignalData'[TYPE]="Poll",'SignalData'[DATE_TIME]>[Past Half Day]),"Past Half Day","")

Past Half Day is a measure, with the simple formula:

Past Half Day = NOW()-(1/2)

The 'Count of Name' column is generated by entering the 'Name' column in my dataset to the Values field and selecting 'Count'. What's really throwing me off is that the individual counts are all correct, it's just the total that is off somehow. The fact that it's reporting 28 consistently tells me that it must be totalling something other than the Count of Name, but I'm not sure what that something is.

Any help or guidance you all can provide would be greatly appreciated.

Regards,

Dan

Re: Filtering to Latest Time Stamp

nickelcap — Thu, 21 Jul 2016 14:56:16 GMT

Hi Herbert,

Yes, I am using the latest version of Power BI Desktop. I made sure that I had the same syntax you showed, and realized that for the last column reference I had managed to select the wrong column! My mistake. The formula is now working as desired.

Thank you very much!

Dan

Re: Filtering to Latest Time Stamp

nickelcap — Fri, 15 Jul 2016 14:16:46 GMT

Hi Herbert,

Thank you very much for the suggestion. I attempted to implement it, however I did not seem to quite get the output I was hoping for. Here's a snip of what I'm looking at:

You can see that for 00011117 the 'Group Latest' is for an earlier time stamp (2016/07/15 2:17:38 AM) than the true latest time stamp (2016/07/15 6:17:36 AM). Looking at my other report, it looks like the time stamps I was working with only had the date. Could adding time to the mix be causing the issue?

Thanks,

Dan

Filtering to Latest Time Stamp

nickelcap — Thu, 14 Jul 2016 16:05:06 GMT

Hi Everyone,

I'm having an issue with filtering by the latest time stamp in Power BI Desktop. I used the method outlined by Lukasz Pawlowski here: https://blogs.msdn.microsoft.com/lukaszp/2015/08/08/finding-the-latest-date-in-power-bi-desktop/.

I've used this method in the past on a different dataset and it works exactly as I want it to. However, when I tried to implement it with a new dataset I found that instead of finding the latest for each category it found the latest for the entire date and time column. I checked and all the syntax is the exact same, aside from me changing the table and column references.

I suspect that the issue is the structure of my data, but I can't figure out what it is. Strangely, if I make a table with the MaxDate measure and one of the other category columns (ex. name) the latest dates for each name are shown. So, I believe that that calculation is good.

Any suggestions on how to troubleshoot this? I'm kind of at a loss since the exact same method worked fine for me before.

Thanks,

Dan

Slicer Breaks Colour Formatting

nickelcap — Thu, 12 May 2016 14:21:00 GMT

Hello,

I created a chart that has some colour formatting set up with it. If a value exceeds 96, the bar is meant to turn red. This worked fine for me until I implemented a slicer that is meant to filter the report down to only values that are below this 96 threshold. When I do this, bars below a value of 96 are red.

The slicer is based off a column which is an if statement based off another column. It reads as: "IF(Appended[Hours Since Last Upload]>96, "Over Time Limit", "Within Normal Time Span")". The chart plots the max of 'Hours Since Last Upload', and colour saturation is set to be based off of 'Hours Since Last Upload'. The Axis is based on a hierarchy, with Customer above Asset, so that the user can drill down from one to the other.

I have Data colours set to Diverging Off, Minimum set to 95, Center is not set as anything, and Maximum set to 96. Minimum is green, center is white, and maximum is red. As I said, this works fine until I apply the 'Within Normal Time Span' side of my slicer. However, when I drill down to 'Asset' on the chart the problem goes away. See the images below:

Is this a bug with Power BI, or have I done something incorrectly?

Thanks for any help you can provide.

Regards,

Dan

Re: Map Visualization Will Not Accept Latitude and Longitude

nickelcap — Tue, 25 Aug 2015 16:01:42 GMT

Worked like a charm, thanks.

Map Visualization Will Not Accept Latitude and Longitude

nickelcap — Tue, 25 Aug 2015 15:22:14 GMT

Hi Everyone,

I've set up a large volume of GPS latitude and longitude values that coincide with time stamps and an RPM value in a table, which I thought could be used in Power BI Desktop to create an interactive map. However, when I try to make the visualization in Power BI Desktop I get the following: Report Pane. When I cursor over the yellow alert icon I get the message: "Location is required to create a map. Click to learn more." Which, when clicked through, does not provide much helpful information.

So far I realized that I needed to set the Long and Lat columns data categories to the correct type, so I've done that. I thought that maybe I had duplicate Long and Lat values so I brought the table back into Excel, removed duplicates from both columns (which really reduced the number of data points, unfortunately).

Any ideas? What am I doing wrong?

Thanks,

Dan