#ask-questions
Title
s
sparse-thailand-8230
10/26/2021, 12:07 PMHello,
Some more questions popped up.
Please help resolve them 🙏
1 . Even if I created 2 variations , The result in both the variations are same. ( Even if i change the value of one variation ,it is affecting the graph on both the variations.) Can we get separate values on variations?
2 . It may have been assigning outliers on its own.
( Getting a wide range of values showing suspicious data, I know it can be resolved by putting capped value.)Is there any way to let it show even if there are outliers ??
( for the suspicious data , a normal warning would be suffice.)
3 . Are there any other significant reports on the results section apart from which is provided?? ( Like graphs or any other charts)
f
future-teacher-7046
10/26/2021, 12:42 PM
1. I'm not sure what you mean. Do you have a screenshot you can share?
2. If you don't set a cap for a metric, no outliers will be removed. We currently don't have any checks for suspicious outliers, but we do plan to add this in the future.
3. In addition to the default report, you can break down results by dimension. We include a built-in Date dimension with graphs that shows results over time. You can add your own custom dimensions as well to break down results by other attributes.
s
sparse-thailand-8230
10/26/2021, 12:44 PMScreenShot of the results with 2 variations.
As for the 3rd question: Are you planning to add any other graphs or anything which can give more insights on the results . Like a cartesain graph which can show the fluctuations of value wrt time.
f
future-teacher-7046
10/26/2021, 1:47 PM
For the experiment with variations showing up the same, can you go to the info tab, click the edit button, and take a screenshot of what the variations section looks like? I have a feeling they both have the same variation_id
We're open to ideas for more graphs to add. Do you have an example you can share of what you're thinking of?
s
sparse-thailand-8230
10/27/2021, 4:08 AMScreenshot of info in experiment.
And for the ideas on result section. I had one which i shared and the bulit in date dimension looks good. Can you please share your "to-add" list , so that we get a better idea on that. Thank you!!!
+ a report from the built-in "date" dimension, Total changes : (Before the experiment and after /
daily/ weeakly/ monthly) for each metric.
ex:
View duration : + 2.45%. (increased)
Revenue: -0.17% (decreased)
Like this
@future-teacher-7046 Also there is an error in Experiment->results->configure Analysis.( " An unexpected error has occurred.")
@future-teacher-7046 @fresh-football-47124 Please Look into it. 🙏
f
fresh-football-47124
10/28/2021, 4:30 AM
Hi Abinash - sorry, didn't see this
s
sparse-thailand-8230
10/28/2021, 4:30 AMNot a problem. 🙂
f
fresh-football-47124
10/28/2021, 4:31 AM
"Can you please share your "to-add" list , so that we get a better idea on that. Thank you!!!" can you explain that a bit more?
s
sparse-thailand-8230
10/28/2021, 4:31 AMBut Could you please go through the previous comments and problem , And provide me some insights on them. 🙏
f
fresh-football-47124
10/28/2021, 4:32 AM
okay, looking, one moment
Where are you seeing the error? On the results page?
Does the developer console give you a better error other than "An unexpected error has occurred"
s
sparse-thailand-8230
10/28/2021, 4:36 AMSure, Because we are dealing with massive data and variations over users in 1.age,2.gender,3.region,4.language etc combinations. Sometimes some of our experiments work well with some groups or sub-groups(let's say a certain combination of the above 4). And it is hard to get insights on which group is performing bad or good or even exceptionally good/bad. So we were expecting the tool to actually point that out so we can work on to solve that problem.
f
fresh-football-47124
10/28/2021, 4:40 AM
okay, ya, the automatic surfacing of interesting dimensions is on our roadmap
s
sparse-thailand-8230
10/28/2021, 4:40 AMIt seems like i am not getting any error in logs.
in Experiment->results->configure Analysis
f
fresh-football-47124
10/28/2021, 4:43 AM
can you press f12 to open the devTools?
see if its throwing an error?
s
sparse-thailand-8230
10/28/2021, 4:45 AMgot it
f
fresh-football-47124
10/28/2021, 4:47 AM
huh
what URL do you get that on?
s
sparse-thailand-8230
10/28/2021, 4:52 AMDo i need to share the url on which i am getting this error?
f
fresh-football-47124
10/28/2021, 4:54 AM
my guess right now is that there is a date column being returned from the db, that our platform is trying to parse and failing. I can add some code to throw a more helpful error, but as we use ISOString code a fair amount, the url would help me track it down. You can DM me if you would rather not share it publicly