Modify index metrics

[Jiaweihu08]

Modify index metrics computation. Some others statistics are added since during many experiments they are found useful.

Changes made:

• Add L1, L2 errors to measure how inner cube sizes deviate from the desiredCubeSize, values are between 0 and 1.
• For inner cubes, add a method to display average normalizedWeight and cube size per level
• Compute depthOnBalance using geometrics sum since many indexes don't have a high number of levels
• Add test for single cube tree

How to use:

import io.qbeast.spark.QbeastTable

val path = "yourTablePath"
val metrics = QbeastTable.forPath(spark, path).getIndexMetrics()

println(metrics)

• Spark Version: 3.1.3
• Hadoop Version: 3.2
• Cluster or local? Local

[osopardo1]

Looks good! The only missing part is to document what are those metrics in the docs/QbeastTable.md file

[codecov]

Codecov Report

Merging #118 (0757c96) into main (5abd0ad) will increase coverage by 1.58%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main     #118      +/-   ##
==========================================
+ Coverage   89.13%   90.71%   +1.58%     
==========================================
  Files          60       60              
  Lines        1307     1325      +18     
  Branches       98      102       +4     
==========================================
+ Hits         1165     1202      +37     
+ Misses        142      123      -19     

Impacted Files	Coverage Δ
src/main/scala/io/qbeast/spark/QbeastTable.scala	88.00% <100.00%> (+37.12%)	⬆️