Radio with custom integer values is not graded properly on correct answer

[MikaelSiidorow]

If I have a submit exercise with a radio field:

# -- snip --
max_points: 1
view_type: access.types.stdsync.createForm

fieldgroups:
  - title: Radio Form
    type: radio
    required: true
    points: 1
    options:
      - label: Option 1
        value: 100
        correct: true
      - label: Option 2
        value: 200
        correct: true

All options are correct so I should receive points regardless of which one I pick, but the result is always 0 points.

I believe I've located the cause to the method grade_radio

mooc-grader/access/types/forms.py

Lines 742 to 759 in 5b8b120

	def grade_radio(self, configuration, value, hints=None):
	hints = hints or []
	# There may be several correct options, but only one of them needs to
	# be selected in the submission in order to gain points.
	correct = False
	i = 0
	for opt in configuration.get("options", []):
	name = self.option_name(i, opt)
	if opt.get("correct", False):
	if name == value:
	correct = True
	else:
	self.append_hint(hints, opt)
	elif name == value:
	self.append_hint(hints, opt)
	i += 1
	return correct, hints, 'string'

Where on line 752 the if name == value is performing a str to int comparison, so it is always false.

[markkuriekkinen]

Thanks for the report! If this is about comparing strings to ints, then what if you write values as strings in the assignment config.yaml file? That is, value: "100"

[MikaelSiidorow]

Thanks for the report! If this is about comparing strings to ints, then what if you write values as strings in the assignment config.yaml file? That is, value: "100"

Yeah, this works perfectly and is the workaround we ended up going with, since there was no need for our data to specifically be integers.