Update modules to ignore bad data

[wbwakeman]

PBS-1868 records whether to "use_spike_data" and "use_lfp_data" for each probe in the release. Need to update LIMS strategies, and SDK modules for each of the following to ignore probes where these fields are FALSE, as noted for each below:

• Current source density
  -- Exclude use_lfp_data = FALSE
• Analysis metrics
  -- Exclude use_spike_data = FALSE
• LFP Subsampling
  -- Exclude use_lfp_data = FALSE

LIMS strategy work tracked in http://jira.corp.alleninstitute.org/browse/PBS-1870

[NileGraddis]

Here is a mapping from bad data interval type -> modules impacted

stimulus

• create_stim_table
• current source density
• write_nwb
• analysis metrics (not yet extant)

spike

• write_nwb
• analysis metrics (not yet extant)

lfp

• current source density
• lfp subsampling
• write nwb

[sgratiy]

@wbwakeman The PR addresses everything except for the Analysis metric module, which does not yet exist.

[sgratiy]

@NileGraddis @wbwakeman listed modules affected by "stimulus". I am not sure what is mean by this category. Also, this looks like a different issue.

[sgratiy]

Associated PR #995