def genRankDragSurvey(surveyname, candidates, shareWith=None):
"""
surveyname (str)
candidates (iterable)
shareWith (str) optional
"""
# connect and craete survey
c = cornellQualtrics()
surveyId = c.createSurvey(surveyname)
desc = "Drag Students into your preferred rank order. Note that the bottom half is undifferentiated."
choices = {}
for j, choice in enumerate(candidates):
choices[str(j + 1)] = {"Display": choice}
choiceOrder = list(range(1, len(choices) + 1))
questionDef = {
"QuestionText": desc,
"DefaultChoices": False,
"DataExportTag": "Q1",
"QuestionType": "RO",
"Selector": "DND",
"SubSelector": "TX",
"Configuration": {
"QuestionDescriptionOption": "UseText"
},
"QuestionDescription": desc,
"Choices": choices,
"ChoiceOrder": choiceOrder,
"Validation": {
"Settings": {
"ForceResponse": "OFF",
"Type": "None"
}
},
"Language": [],
"QuestionID": "QID1",
"QuestionText_Unsafe": desc,
}
c.addSurveyQuestion(surveyId, questionDef)
if shareWith:
c.shareSurvey(surveyId, shareWith)
c.publishSurvey(surveyId)
c.activateSurvey(surveyId)
link = "https://cornell.qualtrics.com/jfe/form/%s" % surveyId
return link
def binRubricSurveyResults(surveyname, outfile):
c = cornellQualtrics()
surveyId = c.getSurveyId(surveyname)
tmpdir = c.exportSurvey(surveyId)
tmpfile = os.path.join(tmpdir, surveyname + ".csv")
assert os.path.isfile(tmpfile), "Survey results not where expected."
res = pandas.read_csv(tmpfile, header=[0, 1, 2])
namecol = res.columns.get_level_values(0)[np.array([
"Select Candidate Name" in c for c in res.columns.get_level_values(1)
])]
names = np.array([n[0] for n in res[namecol].values])
# calculate total scores
quescolinds = np.array([
"Rubric" in c and "Score" in c for c in res.columns.get_level_values(1)
])
quescols = res.columns.get_level_values(0)[quescolinds]
quesnames = res.columns.get_level_values(1)[quescolinds]
scores = res[quescols].values.sum(axis=1)
inds = np.argsort(scores)[::-1]
names = names[inds]
binsize = int(np.round(len(names) * 0.1))
bins = np.zeros(names.size)
for j in range(5):
bins[j * binsize:(j + 1) * binsize] = (j + 1) * 10
bins[bins == 0] = 100
rawscores = res[quescols].values
out = pandas.DataFrame({"Candidate": names, "Bin": bins})
out2 = pandas.DataFrame({
"Candidate": names,
"s1": rawscores[:, 0],
"s2": rawscores[:, 1],
"s3": rawscores[:, 2],
"s4": rawscores[:, 3],
"s5": rawscores[:, 4],
})
with pandas.ExcelWriter(outfile) as ew:
out.to_excel(ew, sheet_name="Reader A Scores", index=False)
out2.to_excel(ew, sheet_name="Reader A Raw Scores", index=False)
def genRankSurvey(readername,
candidates,
binsize,
surveyBaseName=None,
shareWith=None):
"""
readername (str)
candidates (iterable)
binsize (int)
shareWith (str) optional
surveyBaseName (str) optional
"""
# connect and craete survey
c = cornellQualtrics()
surveyname = "Ranking Survey for {}".format(readername)
if surveyBaseName:
surveyname = "{} {}".format(surveyBaseName, surveyname)
surveyId = c.createSurvey(surveyname)
desc = (u"This survey is for: {0}.\n\n"
u"Rank students into the top 50%-ile bins. "
u"Put exactly {1} students in each bin. "
u"All uncategorized students will automatically "
u"be placed in the bottom 50%-ile. Ordering within a bin "
u"does not matter.".format(readername, binsize))
choices = {}
for j, choice in enumerate(candidates):
choices[str(j + 1)] = {"Display": choice}
choiceOrder = list(range(1, len(choices) + 1))
questionDef = {
"QuestionText": desc,
"DefaultChoices": False,
"DataExportTag": "Q1",
"QuestionID": "QID1",
"QuestionType": "PGR",
"Selector": "DragAndDrop",
"SubSelector": "Columns",
"Configuration": {
"QuestionDescriptionOption": "UseText",
"Stack": False,
"StackItemsInGroups": False,
},
"QuestionDescription": desc,
"Choices": choices,
"ChoiceOrder": choiceOrder,
"Validation": {
"Settings": {
"ForceResponse": "ON",
"Type": "GroupChoiceRange",
"MinChoices": "{}".format(binsize),
"MaxChoices": "{}".format(binsize),
}
},
"GradingData": [],
"Language": [],
"NextChoiceId": len(choices) + 1,
"NextAnswerId": 6,
"Groups": ["Top 10%", "Top 20%", "Top 30%", "Top 40%", "Top 50%"],
"NumberOfGroups": 5,
"QuestionText_Unsafe": desc,
}
c.addSurveyQuestion(surveyId, questionDef)
if shareWith:
c.shareSurvey(surveyId, shareWith)
c.publishSurvey(surveyId)
c.activateSurvey(surveyId)
link = "https://cornell.qualtrics.com/jfe/form/%s" % surveyId
return link
def getRankSurveyRes(assignments, outfile, surveyBaseName=None, c=None):
if c is None:
c = cornellQualtrics()
outdict = {}
readersheets = []
for readername in assignments.columns:
surveyname = "Ranking Survey for {}".format(readername)
if surveyBaseName:
surveyname = "{} {}".format(surveyBaseName, surveyname)
surveyId = c.getSurveyId(surveyname)
tmpdir = c.exportSurvey(surveyId)
tmpfile = os.path.join(tmpdir, surveyname + ".csv")
assert os.path.isfile(tmpfile), "Survey results not where expected."
res = pandas.read_csv(tmpfile, header=[0, 1, 2])
if len(res) == 0:
continue
allnames = np.array([])
readernames = []
readerscores = []
for j in range(5):
gcolinds = np.array([
"Q1_{}_GROUP".format(j) in c
for c in res.columns.get_level_values(0)
])
gcols = res.columns.get_level_values(0)[gcolinds]
names = res[gcols].iloc[-1].values
names = names[names == names]
assert len(names) == 3
allnames = np.hstack((allnames, names))
for n in names:
if n in outdict:
outdict[n] += ((j + 1) * 10, )
else:
outdict[n] = ((j + 1) * 10, )
readernames.append(n)
readerscores.append((j + 1) * 10)
unranked = np.array(
list(set(assignments[readername].values) - set(allnames)))
unranked = unranked[unranked != "nan"]
for n in unranked:
if n in outdict:
outdict[n] += (100, )
else:
outdict[n] = (100, )
readernames.append(n)
readerscores.append(100)
readersheets.append(
pandas.DataFrame({
"Name": readernames,
"Rank": readerscores
}))
# build output
outnames = []
outrank1 = []
outrank2 = []
for key, val in outdict.items():
outnames.append(key.split(", "))
outrank1.append(val[0])
if len(val) == 2:
outrank2.append(val[1])
else:
outrank2.append(np.nan)
outnames = np.array(outnames)
out = pandas.DataFrame({
"Last Name": outnames[:, 0],
"First Name": outnames[:, 1],
"Rank 1": outrank1,
"Rank 2": outrank2,
})
with pandas.ExcelWriter(outfile) as ew:
out.to_excel(ew, sheet_name="Ranks", index=False)
for j, r in enumerate(readersheets):
r.to_excel(ew, sheet_name=assignments.columns[j], index=False)
def genRubricSurvey(surveyname,
candidates,
rubrics,
scoreOptions,
shareWith=None):
"""
surveyname (str)
candidates (iterable)
rubrics (iterable)
scoreOptions (iterable)
shareWith (str) optional
"""
# connect and craete survey
c = cornellQualtrics()
surveyId = c.createSurvey(surveyname)
# candidate dropdown
desc = "Select Candidate Name"
choices = {}
for j, choice in enumerate(candidates):
choices[str(j + 1)] = {"Display": choice}
choiceOrder = list(range(1, len(choices) + 1))
questionDef = {
"QuestionText": desc,
"DefaultChoices": False,
"DataExportTag": "Q1",
"QuestionType": "MC",
"Selector": "DL",
"Configuration": {
"QuestionDescriptionOption": "UseText"
},
"QuestionDescription": desc,
"Choices": choices,
"ChoiceOrder": choiceOrder,
"Validation": {
"Settings": {
"ForceResponse": "ON",
"ForceResponseType": "ON",
"Type": "None",
}
},
"Language": [],
"QuestionID": "QID1",
"QuestionText_Unsafe": desc,
}
qid1 = c.addSurveyQuestion(surveyId, questionDef)
questionDef = {
"QuestionText": surveyname,
"DefaultChoices": False,
"DataExportTag": "Q0",
"QuestionType": "DB",
"Selector": "TB",
"Configuration": {
"QuestionDescriptionOption": "UseText"
},
"QuestionDescription": surveyname,
"ChoiceOrder": [],
"Validation": {
"Settings": {
"Type": "None"
}
},
"GradingData": [],
"Language": [],
"NextChoiceId": 4,
"NextAnswerId": 1,
"QuestionID": "QID0",
"QuestionText_Unsafe": surveyname,
}
_ = c.addSurveyQuestion(surveyId, questionDef)
# rubric multiple choice
choices = {}
for j, choice in enumerate(scoreOptions):
choices[str(j + 1)] = {"Display": str(choice)}
choiceOrder = list(range(1, len(choices) + 1))
for j in range(1, len(rubrics) + 1):
desc = rubrics[j - 1]
questionDef = {
"QuestionText": desc,
"DataExportTag": "Q%d" % (j + 1),
"QuestionType": "MC",
"Selector": "SAVR",
"SubSelector": "TX",
"Configuration": {
"QuestionDescriptionOption": "UseText"
},
"QuestionDescription": desc,
"Choices": choices,
"ChoiceOrder": choiceOrder,
"Validation": {
"Settings": {
"ForceResponse": "ON",
"ForceResponseType": "ON",
"Type": "None",
}
},
"Language": [],
"QuestionID": "QID%d" % (j + 1),
"DataVisibility": {
"Private": False,
"Hidden": False
},
"QuestionText_Unsafe": desc,
}
c.addSurveyQuestion(surveyId, questionDef)
questionDef = {
"QuestionText": "Comments",
"DefaultChoices": False,
"DataExportTag": "Q%d" % (len(rubrics) + 2),
"QuestionType": "TE",
"Selector": "SL",
"Configuration": {
"QuestionDescriptionOption": "UseText"
},
"QuestionDescription": "Comments",
"Validation": {
"Settings": {
"ForceResponse": "OFF",
"ForceResponseType": "ON",
"Type": "None",
}
},
"GradingData": [],
"Language": [],
"NextChoiceId": 4,
"NextAnswerId": 1,
"SearchSource": {
"AllowFreeResponse": "false"
},
"QuestionID": "QID%d" % (len(rubrics) + 2),
"QuestionText_Unsafe": "Comments",
}
_ = c.addSurveyQuestion(surveyId, questionDef)
# generate quotas
quotaGroupName = "q1quotas"
quotaGroupId = c.addSurveyQuotaGroup(surveyId, quotaGroupName)
quotas = []
for j, s in enumerate(candidates):
quotaDef = {
"Name": "name{}quota".format(j + 1),
"Occurrences": 1,
"Logic": {
"0": {
"0": {
"LogicType":
"Question",
"QuestionID":
"QID1",
"QuestionIsInLoop":
"no",
"ChoiceLocator":
"q://QID1/SelectableChoice/{}".format(j + 1),
"Operator":
"Selected",
"QuestionIDFromLocator":
"QID1",
"LeftOperand":
"q://QID1/SelectableChoice/{}".format(j + 1),
"Type":
"Expression",
"Description":
"",
},
"Type": "If",
},
"Type": "BooleanExpression",
},
"LogicType": "Simple",
"QuotaAction": "ForBranching",
"ActionInfo": {
"0": {
"0": {
"ActionType": "ForBranching",
"Type": "Expression",
"LogicType": "QuotaAction",
},
"Type": "If",
},
"Type": "BooleanExpression",
},
"QuotaRealm": "Survey",
"Count": 0,
}
quotas.append(c.addSurveyQuota(surveyId, quotaDef))
# and now we can update Q1 with the quotas
desc = "Select Candidate Name"
choices = {}
for j, choice in enumerate(candidates):
choices[str(j + 1)] = {
"Display": choice,
"DisplayLogic": {
"0": {
"0": {
"LogicType": "Quota",
"QuotaID": quotas[j],
"QuotaType": "Simple",
"Operator": "QuotaNotMet",
"LeftOperand": "qo://{}/QuotaNotMet".format(quotas[j]),
"QuotaName": "name{}quota".format(j + 1),
"Type": "Expression",
"Description": "",
},
"Type": "If",
},
"Type": "BooleanExpression",
"inPage": False,
},
}
choiceOrder = list(range(1, len(choices) + 1))
questionDef = {
"QuestionText": desc,
"DefaultChoices": False,
"DataExportTag": "Q1",
"QuestionType": "MC",
"Selector": "DL",
"Configuration": {
"QuestionDescriptionOption": "UseText"
},
"QuestionDescription": desc,
"Choices": choices,
"ChoiceOrder": choiceOrder,
"Validation": {
"Settings": {
"ForceResponse": "ON",
"ForceResponseType": "ON",
"Type": "None",
}
},
"Language": [],
"QuestionID": "QID1",
"QuestionText_Unsafe": desc,
}
c.updateSurveyQuestion(surveyId, qid1, questionDef)
if shareWith:
c.shareSurvey(surveyId, shareWith)
# publish & activate
c.publishSurvey(surveyId)
c.activateSurvey(surveyId)
link = "https://cornell.qualtrics.com/jfe/form/%s" % surveyId
return link