Updates to tripmat and associated functions

[richardellison]

I have updated the tripmat() function (as well as associated functions) to use RSQLite with dbBind instead of dplyr to use COUNT(*) and GROUP_BY(). I have a made a few key changes to the behaviour of tripmat:

1. When trips are filtered by a particular date or time (or both), any trips that have at least a portion fall within the specified dates/times are counted. This is a change from the previous behaviour that required the full trip to be within the specified date ranges.
2. A time of 23 is now interpreted as 23:00:00 to be consistent with earlier times. 24 is interpreted as 23:59:59.
3. tripmat always returns a full matrix, that contains all stations regardless of if trips are used. This compares to how a square matrix was returned but only for stations where at least one trip starts or ends.

[richardellison]

I should add that because of these changes I have updated the tests to ensure they are correct.

[codecov-io]

Codecov Report

❗ No coverage uploaded for pull request base (master@e96aaf9). Click here to learn what that means.
The diff coverage is 91.07%.

@@            Coverage Diff            @@
##             master       #7   +/-   ##
=========================================
  Coverage          ?   83.95%           
=========================================
  Files             ?        4           
  Lines             ?      405           
  Branches          ?        0           
=========================================
  Hits              ?      340           
  Misses            ?       65           
  Partials          ?        0

Impacted Files	Coverage Δ
R/dl_bikedata.R	93.1% <100%> (ø)
src/spatialitedb.cpp	77.5% <100%> (ø)
R/tripmat.R	87.93% <90.38%> (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 e96aaf9...aeb3f5e. Read the comment docs.

[mpadge]

That's great Richard - and you even went beyond the call and modified the tests accordingly! Thanks for the 24->23:59:59 - that had occurred to me too, but only after I'd already committed 23. But mostly, thanks for the excellent index creation function with full >=3.9.0 check - it took me a little while to get my head around what you were doing in create_db_indexes, but it seems like a grand and elegant solution. The filter_datetime with RSQLite is also a great improvement on the previous version.

Now i've gotta get cracking on getting the rest of the data in to the db ...