Looking for a way produce a scatter plot that allows infinite dots (limited potentially only by memory).
The use case it to produce a scatter plot with dots of a settable size (or auto-adjusting based on scatter density) which visually demonstrates each and every single event result from a search.
More-specifically, I want every sign-in to our system to be represented on a scatter plot where X is the time and Y is the service-response-time. With thousands of sign-in's per minute, I need thousands of dots - either 1x1 pixel or 2x2, which results in a cloud-like pattern to be produced along the X series.
when response times go up, the 'cloud' of dots trends upwards and very-clearly identifies when an issue is affecting more than simply outliers.
@dijikul have you tried Splunk Machine Learning Toolkit app from Splunkbase which has several custom visualizations including Scatter Line Chart.
The line scatter plot looks like what I'm envisioning in my head, but I want the X-axis to be time and the Y-axis to be response time. For every sign-in event.
Using an inputlookup in an example is a great way to obfuscate the data so that it's harder to recreate. I have response times; I have a timeline; why does my scatter line plot look like this:
What am I doing wrong?