Moved the normalize_values function to separate file corr_result_helpers.py

Added docstring test to normalize_values

Number of overlapping samples column now displays correctly in the
correlation results page

3 files changed, 44 insertions, 176 deletions

diff --git a/wqflask/base/trait.py b/wqflask/base/trait.py
index 801d32c2..3429d9c1 100755
--- a/wqflask/base/trait.py
+++ b/wqflask/base/trait.py
@@ -286,7 +286,6 @@ class GeneralTrait(object):
                            escape(self.dataset.name),
                            escape(self.name))
             trait_info = g.db.execute(query).fetchone()
-            #print("trait_info is: ", pf(trait_info))
         #XZ, 05/08/2009: We also should use Geno.Id to find marker instead of just using Geno.Name
         # to avoid the problem of same marker name from different species.
         elif self.dataset.type == 'Geno':
@@ -319,7 +318,6 @@ class GeneralTrait(object):
 
             #XZ: assign SQL query result to trait attributes.
             for i, field in enumerate(self.dataset.display_fields):
-                print("  mike: {} -> {} - {}".format(field, type(trait_info[i]), trait_info[i]))
                 setattr(self, field, trait_info[i])
 
             if self.dataset.type == 'Publish':
@@ -329,9 +327,6 @@ class GeneralTrait(object):
 
             self.homologeneid = None
             
-            print("self.geneid is:", self.geneid)
-            print("  type:", type(self.geneid))
-            print("self.dataset.group.name is:", self.dataset.group.name)
             if self.dataset.type == 'ProbeSet' and self.dataset.group and self.geneid:
                 #XZ, 05/26/2010: From time to time, this query get error message because some geneid values in database are not number.
                 #XZ: So I have to test if geneid is number before execute the query.
@@ -356,7 +351,6 @@ class GeneralTrait(object):
                                 InbredSet.SpeciesId = Species.Id AND
                                 Species.TaxonomyId = Homologene.TaxonomyId
                         """ % (escape(str(self.geneid)), escape(self.dataset.group.name))
-                print("-> query is:", query)
                 result = g.db.execute(query).fetchone()
                 #else:
                 #    result = None
@@ -388,7 +382,6 @@ class GeneralTrait(object):
                                 Geno.Name = '{}' and
                                 Geno.SpeciesId = Species.Id
                                 """.format(self.dataset.group.species, self.locus)
-                            print("query is:", query)
                             result = g.db.execute(query).fetchone()
                             self.locus_chr = result[0]
                             self.locus_mb = result[1]
diff --git a/wqflask/utility/corr_result_helpers.py b/wqflask/utility/corr_result_helpers.py
new file mode 100644
index 00000000..edf32449
--- /dev/null
+++ b/wqflask/utility/corr_result_helpers.py
@@ -0,0 +1,30 @@
+def normalize_values(a_values, b_values):
+    """
+    Trim two lists of values to contain only the values they both share
+    
+    Given two lists of sample values, trim each list so that it contains
+    only the samples that contain a value in both lists. Also returns
+    the number of such samples.
+    
+    >>> normalize_values([2.3, None, None, 3.2, 4.1, 5], [3.4, 7.2, 1.3, None, 6.2, 4.1])
+    ([2.3, 4.1, 5], [3.4, 6.2, 4.1], 3)
+    
+    """
+    
+    min_length = min(len(a_values), len(b_values))
+    a_new = []
+    b_new = []
+    for counter in range(min_length):
+        if a_values[counter] and b_values[counter]:
+            a_new.append(a_values[counter])
+            b_new.append(b_values[counter])
+        
+    num_overlap = len(a_new)
+    assert num_overlap == len(b_new), "Lengths should be the same"
+    
+    return a_new, b_new, num_overlap
+
+
+if __name__ == '__main__':
+    import doctest
+    doctest.testmod()
\ No newline at end of file
diff --git a/wqflask/wqflask/correlation/show_corr_results.py b/wqflask/wqflask/correlation/show_corr_results.py
index 3b1ac87d..1410ae0c 100644
--- a/wqflask/wqflask/correlation/show_corr_results.py
+++ b/wqflask/wqflask/correlation/show_corr_results.py
@@ -20,6 +20,9 @@
 
 from __future__ import absolute_import, print_function, division
 
+import sys
+sys.path.append(".")
+
 import gc
 import string
 import cPickle
@@ -43,7 +46,7 @@ from utility.TDCell import TDCell
 from base.trait import GeneralTrait
 from base import data_set
 from base.templatePage import templatePage
-from utility import webqtlUtil, helper_functions
+from utility import webqtlUtil, helper_functions, corr_result_helpers
 from dbFunction import webqtlDatabaseFunction
 import utility.webqtlUtil #this is for parallel computing only.
 from wqflask.correlation import correlationFunction
@@ -122,22 +125,24 @@ class CorrelationResults(object):
 
             self.correlation_data = {}
             for trait, values in self.target_dataset.trait_data.iteritems():
-                this_trait_values = []
-                target_values = []
+                this_trait_vals = []
+                target_vals = []
                 for index, sample in enumerate(self.target_dataset.samplelist):
                     if sample in self.sample_data:
                         sample_value = self.sample_data[sample]
                         target_sample_value = values[index]
-                        this_trait_values.append(sample_value)
-                        target_values.append(target_sample_value)
+                        this_trait_vals.append(sample_value)
+                        target_vals.append(target_sample_value)
+
+                this_trait_vals, target_vals, num_overlap = corr_result_helpers.normalize_values(
+                    this_trait_vals, target_vals)
 
-                this_trait_values, target_values, num_overlap = normalize_values(this_trait_values,
-                                                                                 target_values)
+                print("num_overlap:", num_overlap)
 
                 if self.corr_method == 'pearson':
-                    sample_r, sample_p = scipy.stats.pearsonr(this_trait_values, target_values)
+                    sample_r, sample_p = scipy.stats.pearsonr(this_trait_vals, target_vals)
                 else:
-                    sample_r, sample_p = scipy.stats.spearmanr(this_trait_values, target_values)
+                    sample_r, sample_p = scipy.stats.spearmanr(this_trait_vals, target_vals)
 
                 self.correlation_data[trait] = [sample_r, sample_p, num_overlap]
 
@@ -940,163 +945,3 @@ class CorrelationResults(object):
 
         return traitList
 
-
-    def createExcelFileWithTitleAndFooter(self, workbook=None, identification=None, db=None, returnNumber=None):
-
-        worksheet = workbook.add_worksheet()
-
-        titleStyle = workbook.add_format(align = 'left', bold = 0, size=14, border = 1, border_color="gray")
-
-        ##Write title Info
-        # Modified by Hongqiang Li
-        worksheet.write([1, 0], "Citations: Please see %s/reference.html" % webqtlConfig.PORTADDR, titleStyle)
-        worksheet.write([1, 0], "Citations: Please see %s/reference.html" % webqtlConfig.PORTADDR, titleStyle)
-        worksheet.write([2, 0], "Trait : %s" % identification, titleStyle)
-        worksheet.write([3, 0], "Database : %s" % db.fullname, titleStyle)
-        worksheet.write([4, 0], "Date : %s" % time.strftime("%B %d, %Y", time.gmtime()), titleStyle)
-        worksheet.write([5, 0], "Time : %s GMT" % time.strftime("%H:%M ", time.gmtime()), titleStyle)
-        worksheet.write([6, 0], "Status of data ownership: Possibly unpublished data; please see %s/statusandContact.html for details on sources, ownership, and usage of these data." % webqtlConfig.PORTADDR, titleStyle)
-        #Write footer info
-        worksheet.write([9 + returnNumber, 0], "Funding for The GeneNetwork: NIAAA (U01AA13499, U24AA13513), NIDA, NIMH, and NIAAA (P20-DA21131), NCI MMHCC (U01CA105417), and NCRR (U01NR 105417)", titleStyle)
-        worksheet.write([10 + returnNumber, 0], "PLEASE RETAIN DATA SOURCE INFORMATION WHENEVER POSSIBLE", titleStyle)
-
-        return worksheet
-
-
-    def getTableHeaderForGeno(self, method=None, worksheet=None, newrow=None, headingStyle=None):
-
-        tblobj_header = []
-
-        if method in ["1","3","4"]:
-            tblobj_header = [[THCell(HT.TD(' ', Class="fs13 fwb ffl b1 cw cbrb"), sort=0),
-                              THCell(HT.TD('Record', HT.BR(), 'ID', HT.BR(), Class="fs13 fwb ffl b1 cw cbrb"), text='Record ID', idx=1),
-                              THCell(HT.TD('Location', HT.BR(), 'Chr and Mb', HT.BR(), Class="fs13 fwb ffl b1 cw cbrb"), text='Location (Chr and Mb)', idx=2),
-                              THCell(HT.TD(HT.Href(
-                                                   text = HT.Span('Sample',HT.BR(), 'r', HT.Sup('  ?', style="color:#f00"),HT.BR(), Class="fs13 fwb ffl cw"),
-                                                   target = '_blank',
-                                                   url = "/correlationAnnotation.html#genetic_r"),
-                                           Class="fs13 fwb ffl b1 cw cbrb", nowrap='ON'), text="Sample r", idx=3),
-                              THCell(HT.TD('N',HT.BR(),'Cases',HT.BR(), Class="fs13 fwb ffl b1 cw cbrb"), text="N Cases", idx=4),
-                              THCell(HT.TD(HT.Href(
-                                                   text = HT.Span('Sample',HT.BR(), 'p(r)', HT.Sup('  ?', style="color:#f00"),HT.BR(), Class="fs13 fwb ffl cw"),
-                                                   target = '_blank',
-                                                   url = "/correlationAnnotation.html#genetic_p_r"),
-                                           Class="fs13 fwb ffl b1 cw cbrb", nowrap='ON'), text="Sample p(r)", idx=5)]]
-
-            for ncol, item in enumerate(['Record ID', 'Location (Chr, Mb)', 'Sample r', 'N Cases', 'Sample p(r)']):
-                worksheet.write([newrow, ncol], item, headingStyle)
-                worksheet.set_column([ncol, ncol], 2*len(item))
-        else:
-            tblobj_header = [[THCell(HT.TD(' ', Class="fs13 fwb ffl b1 cw cbrb"), sort=0),
-                              THCell(HT.TD('Record', HT.BR(), 'ID', HT.BR(), Class="fs13 fwb ffl b1 cw cbrb"), text='Record ID', idx=1),
-                              THCell(HT.TD('Location', HT.BR(), 'Chr and Mb', HT.BR(), Class="fs13 fwb ffl b1 cw cbrb"), text='Location (Chr and Mb)', idx=2),
-                              THCell(HT.TD(HT.Href(
-                                                   text = HT.Span('Sample',HT.BR(), 'rho', HT.Sup('  ?', style="color:#f00"),HT.BR(), Class="fs13 fwb ffl cw"),
-                                                   target = '_blank',
-                                                   url = "/correlationAnnotation.html#genetic_rho"),
-                                           Class="fs13 fwb ffl b1 cw cbrb", nowrap='ON'), text="Sample rho", idx=3),
-                              THCell(HT.TD('N',HT.BR(),'Cases',HT.BR(), Class="fs13 fwb ffl b1 cw cbrb"), text="N Cases", idx=4),
-                              THCell(HT.TD(HT.Href(
-                                                   text = HT.Span('Sample',HT.BR(), 'p(rho)', HT.Sup('  ?', style="color:#f00"),HT.BR(), Class="fs13 fwb ffl cw"),
-                                                   target = '_blank',
-                                                   url = "/correlationAnnotation.html#genetic_p_rho"),
-                                           Class="fs13 fwb ffl b1 cw cbrb", nowrap='ON'), text="Sample p(rho)", idx=5)]]
-
-            for ncol, item in enumerate(['Record ID', 'Location (Chr, Mb)', 'Sample rho', 'N Cases', 'Sample p(rho)']):
-                worksheet.write([newrow, ncol], item, headingStyle)
-                worksheet.set_column([ncol, ncol], 2*len(item))
-
-
-        return tblobj_header, worksheet
-
-
-    def getTableBodyForGeno(self, traitList, formName=None, worksheet=None, newrow=None, corrScript=None):
-
-        tblobj_body = []
-
-        for thisTrait in traitList:
-            tr = []
-
-            trId = str(thisTrait)
-
-            corrScript.append('corrArray["%s"] = {corr:%1.4f};' % (trId, thisTrait.corr))
-
-            tr.append(TDCell(HT.TD(HT.Input(type="checkbox", Class="checkbox", name="searchResult",value=trId, onClick="highlight(this)"), nowrap="on", Class="fs12 fwn ffl b1 c222"), text=trId))
-
-            tr.append(TDCell(HT.TD(HT.Href(text=thisTrait.name,url="javascript:showTrait('%s', '%s')" % (formName, thisTrait.name), Class="fs12 fwn ffl"),align="left", Class="fs12 fwn ffl b1 c222"), text=thisTrait.name, val=thisTrait.name.upper()))
-
-            #XZ: trait_location_value is used for sorting
-            trait_location_repr = '--'
-            trait_location_value = 1000000
-
-            if thisTrait.chr and thisTrait.mb:
-                try:
-                    trait_location_value = int(thisTrait.chr)*1000 + thisTrait.mb
-                except:
-                    if thisTrait.chr.upper() == 'X':
-                        trait_location_value = 20*1000 + thisTrait.mb
-                    else:
-                        trait_location_value = ord(str(thisTrait.chr).upper()[0])*1000 + thisTrait.mb
-
-                trait_location_repr = 'Chr%s: %.6f' % (thisTrait.chr, float(thisTrait.mb) )
-
-            tr.append(TDCell(HT.TD(trait_location_repr, Class="fs12 fwn b1 c222", nowrap="on"), trait_location_repr, trait_location_value))
-
-
-            repr='%3.3f' % thisTrait.corr
-            tr.append(TDCell(HT.TD(HT.Href(text=repr, url="javascript:showCorrPlot('%s', '%s')" % (formName, thisTrait.name), Class="fs12 fwn ffl"), Class="fs12 fwn ffl b1 c222", nowrap='ON', align='right'),repr,abs(thisTrait.corr)))
-
-            repr = '%d' % thisTrait.nOverlap
-            tr.append(TDCell(HT.TD(repr, Class="fs12 fwn ffl b1 c222",align='right'),repr,thisTrait.nOverlap))
-
-            repr = webqtlUtil.SciFloat(thisTrait.corrPValue)
-            tr.append(TDCell(HT.TD(repr,nowrap='ON', Class="fs12 fwn ffl b1 c222", align='right'),repr,thisTrait.corrPValue))
-
-            tblobj_body.append(tr)
-
-            for ncol, item in enumerate([thisTrait.name, trait_location_repr, thisTrait.corr, thisTrait.nOverlap, thisTrait.corrPValue]):
-                worksheet.write([newrow, ncol], item)
-            newrow += 1
-
-        return tblobj_body, worksheet, corrScript
-
-def normalize_values(values_1, values_2):
-    N = min(len(values_1), len(values_2))
-    X = []
-    Y = []
-    for i in range(N):
-        if values_1[i]!= None and values_2[i]!= None:
-            X.append(values_1[i])
-            Y.append(values_2[i])
-    num_overlap = len(X)
-            
-    return (X, Y, num_overlap)
-
-
-def cal_correlation(values_1, values_2):
-    N = min(len(values_1), len(values_2))
-    X = []
-    Y = []
-    for i in range(N):
-        if values_1[i]!= None and values_2[i]!= None:
-            X.append(values_1[i])
-            Y.append(values_2[i])
-    NN = len(X)
-    if NN <6:
-        return (0.0,NN)
-    sx = reduce(lambda x,y:x+y,X,0.0)
-    sy = reduce(lambda x,y:x+y,Y,0.0)
-    x_mean = sx/NN
-    y_mean = sy/NN
-    xyd = 0.0
-    sxd = 0.0
-    syd = 0.0
-    for i in range(NN):
-        xyd += (X[i] - x_mean)*(Y[i] - y_mean)
-        sxd += (X[i] - x_mean)*(X[i] - x_mean)
-        syd += (Y[i] - y_mean)*(Y[i] - y_mean)
-    try:
-        corr = xyd/(sqrt(sxd)*sqrt(syd))
-    except:
-        corr = 0
-    return (corr, NN)