Add tests for all tasks and workflows

[drmani]

Add tests using data subsets, synthetic data, or full datasets. Each task and workflow should have a test method.

[ramanibk]

Test Script now exists in the PANOPLY/tests/latest folder that compares a gold-standard output tarball with a test tarball.

Function

• Checks for missing png, jpeg, and pdf files in addition to other "data" files
• Checks for missing sub-directories
• Compares all GCTs and CSVs unless specified with other regular expression rules to narrow down the list of files to compare
• Uses a tolerance argument` that can be set while calling the script
• Logs messages for failed comparisons, missing files, and passed files

Details

• Inputs:
  • gold.tar: gold standard tarball
  • test.tar: test tarball
  • tolerance: threhsold for comparing numerical matrices; by default set to 1.490116e-08 or .Machine$double.eps ^ 0.5
  • file.list: a YAML file listing patterns
• More on the YAML file:
  • keys should be sub-directories inside the tarball (NOTE: if sub-directory name is harmonized-data use key make.names( "harmonized-data" ) as the key, which in this case would be harmonized.data
  • each sub-directory should have a non-nested list of patterns
  • patterns should be regular expressions for file matching with list.files()
  • patterns can also be of the form "!<pattern>" (ex. "!sample-info.csv" or "!*.tsv" )
  • by default, this script will compare all CSV files assuming that first row will be its header
  • by default, this script will compare all GCT files' @mat attributes
• Report will be stored as test-report.txt

NOTE that, this script does one test for all the sub-directories. If you want to test for just one module, this should still work as long as you give it correct inputs. Other directories will simply be listed as missing.