fix indoor_eval in case of less classes in prediction

[filaPro]

Fixes #230.

[CLAassistant]

All committers have signed the CLA.

[codecov]

Codecov Report

Merging #231 (d4f186e) into master (bafee2e) will not change coverage.
The diff coverage is 0.00%.

@@           Coverage Diff           @@
##           master     #231   +/-   ##
=======================================
  Coverage   49.94%   49.94%           
=======================================
  Files         172      172           
  Lines       11558    11558           
  Branches     1801     1801           
=======================================
  Hits         5773     5773           
  Misses       5411     5411           
  Partials      374      374           

Flag	Coverage Δ
unittests	49.94% <0.00%> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
mmdet3d/core/evaluation/indoor_eval.py	10.67% <0.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update bafee2e...caf24ba. Read the comment docs.

[Tai-Wang]

Please notice messages from linting. You can pre-commit to check whether there is any problem in terms of format.

[filaPro]

Hi @Tai-Wang,

I've fixed yapf warnings. However I have one more question. What is the idea behind origin=(0.5, 0.5, 0.5) here? Looks like this function deals with gt and pred in terms of DepthInstance3DBoxes with default origin == (0.5, 0.5, 0.0). But in this line you only shift gt along z axis and don't modify z coordinate of pred. Is this the expected behavour?