def test1(self):
"""1 way anova"""
R="""Anova: Single Factor on Measure
SUMMARY
Groups Count Sum Average Variance
==========================================
A 14 870 62.143 202.132
B 15 858 57.200 584.743
C 16 1510 94.375 339.583
ANOVA
Source of SS df MS F P-value
Variation
===========================================================
Treatments 12656.047 2 6328.023 16.707 4.589e-06
Error 15907.864 42 378.759
===========================================================
Total 28563.911 44 """
listOflists=[[42,52,55,59,75,40,79,79,44,56,68,77,75,69],
[29,36,29,31,97,88,27,57,54,77,54,52,58,91,78],
[91,79,73,75,99,66,114,120,102,68,114,79,115,104,107,104]]
D=Anova1way()
D.run(listOflists)
self.assertEqual(str(D),R)
def test0(self):
# http://www.utdallas.edu/~herve/abdi-NewmanKeuls2010-pretty.pdf
d = [[21.0, 20.0, 26.0, 46.0, 35.0, 13.0, 41.0, 30.0, 42.0, 26.0],
[23.0, 30.0, 34.0, 51.0, 20.0, 38.0, 34.0, 44.0, 41.0, 35.0],
[35.0, 35.0, 52.0, 29.0, 54.0, 32.0, 30.0, 42.0, 50.0, 21.0],
[44.0, 40.0, 33.0, 45.0, 45.0, 30.0, 46.0, 34.0, 49.0, 44.0],
[39.0, 44.0, 51.0, 47.0, 50.0, 45.0, 39.0, 51.0, 39.0, 55.0]]
conditions_list = 'Contact Hit Bump Collide Smash'.split()
D=Anova1way()
D.run(d, conditions_list=conditions_list)
def test0(self):
"""1 way anova"""
R="Anova1way([('f', 16.70726997413529), ('p', 4.5885798225758395e-06), ('ns', [14, 15, 16]), ('mus', [62.142857142857146, 57.2, 94.375]), ('vars', [202.13186813186815, 584.7428571428571, 339.5833333333333]), ('ssbn', 12656.046825396828), ('sswn', 15907.864285714284), ('dfbn', 2), ('dfwn', 42), ('msbn', 6328.023412698414), ('mswn', 378.7586734693877)], conditions_list=['A', 'B', 'C'])"
listOflists=[[42,52,55,59,75,40,79,79,44,56,68,77,75,69],
[29,36,29,31,97,88,27,57,54,77,54,52,58,91,78],
[91,79,73,75,99,66,114,120,102,68,114,79,115,104,107,104]]
D=Anova1way()
D.run(listOflists)
self.assertEqual(repr(D),R)
def getCatAlphaData(request):
### get sample list from cookie
samples = Sample.objects.all()
samples.query = pickle.loads(request.session['selected_samples'])
selected = samples.values_list('sampleid')
qs1 = Sample.objects.all().filter(sampleid__in=selected)
if request.is_ajax():
allJson = request.GET["all"]
all = simplejson.loads(allJson)
button = int(all["button"])
sig_only = int(all["sig_only"])
norm = int(all["normalize"])
selectAll = int(all["selectAll"])
metaString = all["meta"]
### function to merge values on common keys
metaDict = simplejson.JSONDecoder(
object_pairs_hook=multidict).decode(metaString)
### function to create a meta variable DataFrame
metaDF = catAlphaMetaDF(qs1, metaDict)
myList = metaDF['sampleid'].tolist()
mySet = list(set(myList))
### function to create a taxa DataFrame
taxaDF = taxaProfileDF(mySet)
### function to merge values on common keys
taxaString = all["taxa"]
### this taxaDict is from the dynatree (ajax call)
taxaDict = simplejson.JSONDecoder(
object_pairs_hook=multidict).decode(taxaString)
# change dict if selectAll levels is on (avoids loading entire tree first)
if selectAll == 1:
taxaDict = {}
qs1 = Profile.objects.all().filter(sampleid__in=mySet).values_list(
'kingdomid', flat='True').distinct()
taxaDict['Kingdom'] = qs1
elif selectAll == 2:
taxaDict = {}
qs1 = Profile.objects.all().filter(sampleid__in=mySet).values_list(
'phylaid', flat='True').distinct()
taxaDict['Phyla'] = qs1
elif selectAll == 3:
taxaDict = {}
qs1 = Profile.objects.all().filter(sampleid__in=mySet).values_list(
'classid', flat='True').distinct()
taxaDict['Class'] = qs1
elif selectAll == 4:
taxaDict = {}
qs1 = Profile.objects.all().filter(sampleid__in=mySet).values_list(
'orderid', flat='True').distinct()
taxaDict['Order'] = qs1
elif selectAll == 5:
taxaDict = {}
qs1 = Profile.objects.all().filter(sampleid__in=mySet).values_list(
'familyid', flat='True').distinct()
taxaDict['Family'] = qs1
elif selectAll == 6:
taxaDict = {}
qs1 = Profile.objects.all().filter(sampleid__in=mySet).values_list(
'genusid', flat='True').distinct()
taxaDict['Genus'] = qs1
elif selectAll == 7:
taxaDict = {}
qs1 = Profile.objects.all().filter(sampleid__in=mySet).values_list(
'speciesid', flat='True').distinct()
taxaDict['Species'] = qs1
factor = 'none'
if norm == 1:
factor = 'none'
elif norm == 2:
factor = 'min'
elif norm == 3:
factor = '10th percentile'
elif norm == 4:
factor = '25th percentile'
elif norm == 5:
factor = 'median'
### function to normalize the number of sequence reads per sample
normDF = normalizeAlpha(taxaDF, taxaDict, mySet, factor)
finalDF = metaDF.merge(normDF, on='sampleid', how='outer')
finalDF[['count', 'rel_abund', 'rich', 'diversity'
]] = finalDF[['count', 'rel_abund', 'rich',
'diversity']].astype(float)
pd.set_option('display.max_rows', finalDF.shape[0],
'display.max_columns', finalDF.shape[1], 'display.width',
1000)
final_fieldList = []
for key in metaDict:
final_fieldList.append(key)
finalDict = {}
result = ""
seriesList = []
xAxisDict = {}
yAxisDict = {}
### group DataFrame by each taxa level selected
grouped1 = finalDF.groupby(['rank', 'taxa_name', 'taxa_id'])
equal_error = 'no'
### group DataFrame by each meta variable selected
for name1, group1 in grouped1:
trtList = []
valList = []
grouped2 = pd.DataFrame()
if button == 1:
grouped2 = group1.groupby(final_fieldList)['count']
elif button == 2:
grouped2 = group1.groupby(final_fieldList)['rel_abund']
elif button == 3:
grouped2 = group1.groupby(final_fieldList)['rich']
### for taxa with only 1 species all values will be '1' and cause an anova error
if group1['rich'].sum() == group1['rich'].count():
equal_error = 'yes'
elif button == 4:
grouped2 = group1.groupby(final_fieldList)['diversity']
for name2, group2 in grouped2:
if isinstance(name2, unicode):
trt = name2
else:
trt = ' & '.join(list(name2))
trtList.append(trt)
valList.append(list(group2.T))
### One-way ANOVA with some error checking
D = Anova1way()
if equal_error == 'no':
try:
D.run(valList, conditions_list=trtList)
anova_error = 'no'
except:
D['p'] = 1
anova_error = 'yes'
else:
D['p'] = 1
anova_error = 'yes'
### select only significant ANOVAs for output (graph & text area)
if sig_only == 1:
if D['p'] <= 0.05:
result = result + '===============================================\n'
result = result + 'Taxa level: ' + str(name1[0]) + '\n'
result = result + 'Taxa name: ' + str(name1[1]) + '\n'
if button == 1:
result = result + 'Dependent Variable: Sequence Reads' + '\n'
elif button == 2:
result = result + 'Dependent Variable: Relative Abundance' + '\n'
elif button == 3:
result = result + 'Dependent Variable: Species Richness' + '\n'
elif button == 4:
result = result + 'Dependent Variable: Shannon Diversity' + '\n'
indVar = ' x '.join(final_fieldList)
result = result + 'Independent Variable: ' + str(
indVar) + '\n'
if equal_error == 'yes' or anova_error == 'yes':
result = result + 'Analysis cannot be performed...' + '\n'
else:
result = result + str(D) + '\n'
result = result + '===============================================\n'
result = result + '\n\n\n\n'
dataList = []
grouped2 = group1.groupby(final_fieldList).mean()
if button == 1:
dataList.extend(list(grouped2['count'].T))
elif button == 2:
dataList.extend(list(grouped2['rel_abund'].T))
elif button == 3:
dataList.extend(list(grouped2['rich'].T))
elif button == 4:
dataList.extend(list(grouped2['diversity'].T))
seriesDict = {}
seriesDict['name'] = name1
seriesDict['data'] = dataList
seriesList.append(seriesDict)
xTitle = {}
xTitle['text'] = indVar
xAxisDict['title'] = xTitle
xAxisDict['categories'] = trtList
yTitle = {}
if button == 1:
yTitle['text'] = 'Sequence Reads'
elif button == 2:
yTitle['text'] = 'Relative Abundance'
elif button == 3:
yTitle['text'] = 'Species Richness'
elif button == 4:
yTitle['text'] = 'Shannon Diversity'
yAxisDict['title'] = yTitle
### select all ANOVAs for output (graph & text area)
if sig_only == 0:
result = result + '===============================================\n'
result = result + 'Taxa level: ' + str(name1[0]) + '\n'
result = result + 'Taxa name: ' + str(name1[1]) + '\n'
if button == 1:
result = result + 'Dependent Variable: Sequence Reads' + '\n'
elif button == 2:
result = result + 'Dependent Variable: Relative Abundance' + '\n'
elif button == 3:
result = result + 'Dependent Variable: Species Richness' + '\n'
elif button == 4:
result = result + 'Dependent Variable: Shannon Diversity' + '\n'
indVar = ' x '.join(final_fieldList)
result = result + 'Independent Variable: ' + str(indVar) + '\n'
if equal_error == 'yes' or anova_error == 'yes':
result = result + 'Analysis cannot be performed...' + '\n'
else:
result = result + str(D) + '\n'
result = result + '===============================================\n'
result = result + '\n\n\n\n'
dataList = []
grouped2 = group1.groupby(final_fieldList).mean()
if button == 1:
dataList.extend(list(grouped2['count'].T))
elif button == 2:
dataList.extend(list(grouped2['rel_abund'].T))
elif button == 3:
dataList.extend(list(grouped2['rich'].T))
elif button == 4:
dataList.extend(list(grouped2['diversity'].T))
seriesDict = {}
seriesDict['name'] = name1
seriesDict['data'] = dataList
seriesList.append(seriesDict)
xTitle = {}
xTitle['text'] = indVar
xAxisDict['title'] = xTitle
xAxisDict['categories'] = trtList
yTitle = {}
if button == 1:
yTitle['text'] = 'Sequence Reads'
elif button == 2:
yTitle['text'] = 'Relative Abundance'
elif button == 3:
yTitle['text'] = 'Species Richness'
elif button == 4:
yTitle['text'] = 'Shannon Diversity'
yAxisDict['title'] = yTitle
finalDict['series'] = seriesList
finalDict['xAxis'] = xAxisDict
finalDict['yAxis'] = yAxisDict
finalDict['text'] = result
if not seriesList:
finalDict['empty'] = 0
else:
finalDict['empty'] = 1
finalDF.reset_index(drop=True, inplace=True)
res_table = finalDF.to_html(classes="table display")
res_table = res_table.replace('border="1"', 'border="0"')
finalDict['res_table'] = str(res_table)
res = simplejson.dumps(finalDict)
return HttpResponse(res, content_type='application/json')