datalust / seq-tickets

Issues, design discussions and feature roadmap for the Seq log server
https://datalust.co/seq
98 stars 5 forks source link

Analyze distributions with histograms and heatmaps #2303

Open liammclennan opened 4 weeks ago

liammclennan commented 4 weeks ago

A graphical view of the distribution of numbers can help show where most values are clustered and the presence of any high or low outliers. The standard visualization for distributions is the histogram, and the heatmap to view distributions over time (or any other dimension). Each column in a heatmap is equivalent to a histogram.

Add a 'Plot distribution' option to numeric properties to generate a historgram. A histogram can be generated for any query result with two columns if the second column is numeric. Read more about historgrams.

Add a 'Plot distribution over time' option to numeric properties to generate a heatmap. A heatmap can be generated for any query result with three columns if the third column is numeric. Read more about heatmaps.

Histogram

image A histogram showing the distribution of @Elapsed values.

Heatmap

image

image A heatmap of categorical data.