CC
Attachment for the speedo article I just sent
I did not know it needed zipping and instead of asking me for a zipped file it just sent off the message without the attachment
Hi all
I have not heard from anyone on this one yet so combined the earlier text with the attachment in case that was the issue on feedback
I hope someone can help me on this one because I have tried for days now to get to the bottom of why a datafeed that has been previously attached to a table that gives me the right result in the table yet when I take that same datafeed and attach it to a speedo is does not.
I have even gone to great lengths to sort this out and in the end I took the results from my MashApp table to CSV as the datafeed to the table had a significant amount of iterations to arrive at the right result and I wanted to remove this from the equation.
I then imported the already calculated datafeed to MashZone and checked the aggregation and sums in the output and they were spot on - having said that the original datafeed I had when I applied the aggregation only in the output view only calculated the correct result on three out of four rows - yet in the MashApp table it was 100% correct - very concerning.
Parking that for one minute on the clean imported datafeed that I only carried out a simple calculation on the % difference from one column to another (that calculation was even correct when I checked the output) - but when I attached it to a speedo and filtered it on a MashApp by one of the fields using a list box for each of the four rows it produced the correct result for each of the rows (having previously done the calculations in Excel to check what the result should be) - but when unselecting the rows to look at the total percentage it was wrong - it should have been 98.65 yet the speedo returned 88.18
I am at a loss to establish why a speedo does not reflect the results of the datafeed output and why the output in a datafeed does not reflect the results of a table on a MashApp
Producing a result in MashZone is fairly straight forward but producing the right result and being able to easily validate it on a datafeed output is a nightmare
I have attached the simple version final output I exported from the MashApp table that when attached as a new datafeed only had the percentage calculation (Qty-2 minus Qty-1) – if the totals speedo issue can be solved on this then it may also resolve the issue I have on the speedo when looking at the same datafeed that the table used as that had even different numbers on the speed to the cleaned up version - I live in hope
Regards
Charles
Dear Charles,
sorry that i'm the first that gets aware of your post so loang after you wrote it.
Can you please export your MashApp as MZP file via the MashZone Administration and send it to us (or attach it). Otherwise we cannot trace what's happening in your example.
Kind regards
Jürgen Kirsch
Hi Jurgen
Many thanks for getting back to me I was beginning to think my issue was too difficult to solve.
I have exported the MZP file but don’t know how to attach it as there does not appear to be an attach option on the reply to this post or if there is then I dont know how to find it
Regards
Charles
Hi Charles,
i had a look at the MashApp & Feed that you sent me. I cannot tell if the version that you gave me is exactly the same as when you wrote your post. So i will just write what came to my attention when looking at it.
In short: MashZone is right in what it calculates and displays. I think you mixed some things. I wil try to clarify and i hope the calculated results of MashZone will then make sense to you.
Why does the Speedometer show 88.18?
You selected AVG as calculation method for the feed column 'change' that you have connected to the speedometer. That means: All rows of the feed are taken and the calculated average will be displayed as speedometer value. Excel calculates the same value as average. So what you see is the average 'change' over all 'names'.
Please see the following 2 screenshots.
.png)
Maybe you tried to show the change for the currently selected name in the pie chart? To do this you have to define a filter in the data assignmnt mode of the speedometer to use the selected pie slice (the name) to show only the value of the 'change' for the currently selected name.
Further: In your feed you already calculate the 'change' as percentage value between Qty-1 anf Qty-2. Then you display this in the pie chart. There, 2 values are shown in the tooltip of each individual pie slice. First value being the absolute value of 'change' - which is already a percentage value and th second value being a percentage value that is calculated automatically by the pie chart being the percentage of this pie slice in comparison to the other 4 pie slices. I think this percentage value does not make much sense here, since it's the percentage of a percentage KPI - something you cannot evaluate as a person in a meaningful way and it's also easily mistaken with the primarily kpi 'change' itself.
I suggest to use a bar chart instead of a pie to show an individual bar for each name. This way you can separate them from eachother and make the true amount of 'change' visible. Something that is not possible with the pie chart.
Regards
Jürgen Kirsch
