[FIX] Discourage use of "sample" in events if sampling frequency is ambiguous, add guidelines for precision

[sappelhoff]

closes #1043

Adds two minor tweaks to our documentation of events.tsv:

1. Don't Use sample when the sampling frequency in the raw data is ambiguous / mixed
2. Use a reasonable precision for numeric metadata (onset, duration, response_time, ...)

[codecov]

Codecov Report

Merging #1140 (2692bf5) into master (7172921) will increase coverage by 2.90%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #1140      +/-   ##
==========================================
+ Coverage   75.18%   78.08%   +2.90%     
==========================================
  Files          11       11              
  Lines        1362     1378      +16     
==========================================
+ Hits         1024     1076      +52     
+ Misses        338      302      -36     

Impacted Files	Coverage Δ
...ools/schemacode/schemacode/tests/test_validator.py	100.00% <0.00%> (ø)
tools/schemacode/schemacode/_version.py	38.90% <0.00%> (+2.59%)	⬆️
tools/schemacode/schemacode/validator.py	92.13% <0.00%> (+9.13%)	⬆️

---

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 07741e7...2692bf5. Read the comment docs.

[VisLab]

This is good.

Also suggest changing "referred" to "referenced" on line 70.