This PR addresses the issue #36. For the first test, candidates will generate a .yml file with the images that they identified that need correction. This PR adds a script that compares .yml files generated by the candidates and the reference list. Every .yml file from candidates will be in a folder and the script loops through all files and compares with ref.
Outputs log file and a .csv with the results.
Results are:
If the file was unable to be loaded, for the instance FORMATING, ERROR will appear.
If the .yml file doesn't start with "FILESEG:", the instance FILESEG will be set to False.
Number of files identified refers to the total of files that the candidate identified.
Ex: 6/10 indicates that the candidate identified 6 images that need correction when there is acutally 10 to be identified.
Nb_right_files refers to number of images correctly identified.
Ex: 3/6 indicates that out of the 6 images that the candidate identified, only 3 truly needed manual correction.
Wrongly identified files are listed in the log file. (For reference, if all candidates make the same mistakes, ...)
Description
This PR addresses the issue #36. For the first test, candidates will generate a .yml file with the images that they identified that need correction. This PR adds a script that compares .yml files generated by the candidates and the reference list. Every .yml file from candidates will be in a folder and the script loops through all files and compares with ref.
Outputs log file and a .csv with the results.
Results are:
FORMATING
,ERROR
will appear.FILESEG
will be set toFalse
.Number of files identified
refers to the total of files that the candidate identified.Nb_right_files
refers to number of images correctly identified.Example of output
(The last table is also saved in a .csv file)