High density scatter chart based on date

[Anonymous]

I am struggling with a scatter chart where i am based on date and value. 

What I have now is: 

 And what I would love to have is:

 What i tried is: 

- high density line chart, it looks better but you cannot point out to the specific dot, it takes random from area and the dot size is 5 the smallest

- make a test date based on day or month but it takes as a sum of this day/month even though I choose "don`t summarize". 

The date range will be taken from slicer. 

Any ideas? 

Thanks in advance. 

[v-shex-msft]

HI @Anonymous,

Can please share some dummy data with a similar data structure to test? It should help us clarify your scenario and test to design scatter charts.

How to Get Your Question Answered Quickly 

Regards,

Xiaoxin Sheng

[Anonymous]

Hi Xiaoxin, 

I have 2 tables, where I want to connect two columns, one from each table (TestDate and Value). They are connected with ID column. TestDate could be categorized as days or months but when i try to do so, Values are getting sumed up and shows one dot instead of intensed scatter chart. 

I hope it helps somehow to understand what i want to acheive. If not, please let me know. 

[v-shex-msft]

HI @Anonymous,

I test with some dummy data that similar to your snapshot. I test on 'scatter chart' but found it seems only allows you to use the numeric value on it. In addition, it also not support you to directly show values without aggregation groups.

If you want to achieve the high-density chart, you need to use the numeric axis to replace the date value first.
Then you need to expand the table value to add some dummy data based raw values and random offsets to fill the nearest data ranges.

BTW, you can also try to use script-based visuals with the script to manually plot these graphs.

Scatter Plots - R Base Graphs 

Regards,

Xiaoxin Sheng

[Anonymous]

Dear Xiaoxin, 

many thanks for your input. I tried your solution but again with no luck. I tried to replace the date with numeric value however it always sum up the value even I choose don`t summarize. 

I "played" around with different options and what i did, is changed from scatter to line chart and almost achieved what I wanted. 

I see 2 issues now: 

1. dots cannot be lower than size 5, which I have seen on forums it`s still impossible to go lower and it will be a voting done for it

2. Dots cannot be clicked/picked as I want. Once I click the specific dot, it takes some from the area (below I clicked the toppest one and it took just random): 

Do you know any workaround for that? This chart takes data as i wanted based on years (which is fine and I couldnt achieve the same on scatter chart). 

Many thanks for your help. 

[v-shex-msft]

HI @Anonymous,

#1, For the minimum data point size, it may be a limit of the chart render. I'd like to suggest you vote for these ideas to improve the visuals features.
#2, I don't think it is possible to achieve this on a high-density data label chart, it is hard to accurately recognize each data point that your mouse located. (probably due to performance reasons)

For this scenario, power bi may pickup a random data point or show the summarized value of the current axis label instead.

Regards,

Xiaoxin Sheng

[amitchandak]

@Anonymous , the second chart is typically drawn for measure vs measure(power bi standard scatter visual) . In case of the date in on the x-axis , it will kind dot visual(without line) 

You might check for custom visual or python visual

https://appsource.microsoft.com/en-us/marketplace/apps?product=power-bi-visuals