[FIX] Address divide by 0 issue in decay.py

[handwerkerd]

For 5 echo data, I was noticing the following warning:

INFO     tedana:tedana_workflow:584 Computing T2* map
/Users/handwerkerd/code/tedana_community/me-ica/tedana/tedana/decay.py:40: RuntimeWarning: divide by zero encountered in true_divide
  temp_arr = np.exp(-echo_times / t2s)  # (E x V) array
INFO     combine:make_optcom:242 Optimally combining data with voxel-wise T2* estimates

This warning was created by #736 when it became possible to to send a t2s volume with 0's to _apply_t2s_floor The function correctly ignored voxels where t2s was zero, but only after dividing by t2s and triggering this warning

Changes proposed in this pull request:

• Edited _apply_t2s_floor so that voxels with 0 in t2s are excluded from the denominator when dividing

[codecov]

Codecov Report

Merging #807 (b1e1f38) into main (0f04f76) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main     #807   +/-   ##
=======================================
  Coverage   93.26%   93.27%           
=======================================
  Files          27       27           
  Lines        2213     2215    +2     
=======================================
+ Hits         2064     2066    +2     
  Misses        149      149           

Impacted Files	Coverage Δ
tedana/decay.py	97.27% <100.00%> (+0.03%)	⬆️

---

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 0f04f76...b1e1f38. Read the comment docs.

[jbteves]

LGTM, thanks for patching this @handwerkerd !

[jbteves]

@handwerkerd would you mind prepending the [FIX] tag to this PR?

[tsalo]

It looks good to me too. Thanks!