Test dates for matching data with metadata

[peterdesmet]

Data will be matched to receivers using receiver_id and date_time. Errors or missing information in the metadata can cause errors though. Here are a number of test cases and the predicted output. We'll have to review if this is adequate enough.

Receiver only deployed once

Test date: 2012-06-30

• [x] a-single: 2012-07-01 / 2014-07-09 → No match
• [x] b-single: 2012-07-01 / null → No match
• [x] c-single: null / 2014-07-09 → Match before
• [x] d-single: 2014-07-09 / 2012-07-01 (error in metadata) → No match
• [x] e-single: null / null → Match

  Test date: 2013-01-01

• [x] a-single: 2012-07-01 / 2014-07-09 → Match between
• [x] b-single: 2012-07-01 / null → Match after
• [x] c-single: null / 2014-07-09 → Match before
• [x] d-single: 2014-07-09 / 2012-07-01 (error in metadata) → No match
• [x] e-single: null / null → Match

  Test date: 2014-07-10

• [x] a-single: 2012-07-01 / 2014-07-09 → No match
• [x] b-single: 2012-07-01 / null → Match after
• [x] c-single: null / 2014-07-09 → No match
• [x] d-single: 2014-07-09 / 2012-07-01 (error in metadata) → No match
• [x] e-single: null / null → Match

  Receiver deployed two times

  Test date: 2012-06-30

• [x] a-double: 2012-01-01 / 2012-06-29, 2012-07-01 / 2014-07-09 → No match (between pairs)
• [x] b-double: null / 2012-06-29, 2012-07-01 / null → No match
• [x] c-double: null / 2012-07-01, 2012-06-29 / null (error in metadata) → Match for both
• [x] d-double: 2012-01-01 / null, null / 2014-07-09 (missing info in metadata) → Match between both
• [x] e-double: null / 2012-06-29, null / 2014-07-09 (missing info in metadata) → Match before second
• [x] f-double: 2012-01-01 / null, 2012-07-01 / null (missing info in metadata) → Match after first
• [x] g-double: null / null, null / null (missing info in metadata) → Match for both

  Test date: 2013-01-01

• [x] a-double: 2012-01-01 / 2012-06-29, 2012-07-01 / 2014-07-09 → Match between second pair
• [x] b-double: null / 2012-06-29, 2012-07-01 / null → Match after second
• [x] c-double: null / 2012-07-01, 2012-06-29 / null (error in metadata) → Match after second
• [x] d-double: 2012-01-01 / null, null / 2014-07-09 (missing info in metadata) → Match between both
• [x] e-double: null / 2012-06-29, null / 2014-07-09 (missing info in metadata) → Match before second
• [x] f-double: 2012-01-01 / null, 2012-07-01 / null (missing info in metadata) → Match after both
• [x] g-double: null / null, null / null (missing info in metadata) → Match for both

  Test date: 2014-07-10

• [x] a-double: 2012-01-01 / 2012-06-29, 2012-07-01 / 2014-07-09 → No match
• [x] b-double: null / 2012-06-29, 2012-07-01 / null → Match after second
• [x] c-double: null / 2012-07-01, 2012-06-29 / null (error in metadata) → Match after second
• [x] d-double: 2012-01-01 / null, null / 2014-07-09 (missing info in metadata) → Match after first
• [ ] e-double: null / 2012-06-29, null / 2014-07-09 (missing info in metadata) → No match
• [ ] f-double: 2012-01-01 / null, 2012-07-01 / null (missing info in metadata) → Match after both
• [ ] g-double: null / null, null / null (missing info in metadata) → Match for both

[peterdesmet]

I created an SQL querie that passes the test, but we'll have to discuss if those results are what we actually want.