working numpy version

FAVE-extract/bin/extractFormants.py

@@ -71,7 +71,9 @@
 import praat, esps, plotnik, cmu, vowel
 
 FAVE_WD = os.getcwd()
-import rpy2.robjects as robjects
+import numpy as np
+from scipy import linalg
+from scipy.spatial.distance import mahalanobis as mahalanobis
 import remeasure
 os.chdir(FAVE_WD)
 
@@ -1051,9 +1053,9 @@ def loadCovs(inFile):
     covs = {}
     for line in open(inFile, 'rU').readlines():
         vowel = line.strip().split('\t')[0]
-        values = robjects.FloatVector(line.strip().split('\t')[1:])
-        covs[vowel] = robjects.r['matrix'](values, nrow=4)
-        
+        values = np.array([float(x) for x in line.strip().split('\t')[1:]])
+        covs[vowel] = np.reshape(values, (4,-1))        
+
     return covs
 
 
@@ -1063,8 +1065,8 @@ def loadMeans(inFile):
     means = {}
     for line in open(inFile, 'rU').readlines():
         vowel = line.strip().split('\t')[0]
-        means[vowel] = robjects.FloatVector(line.strip().split('\t')[1:])
-        
+        means[vowel] = np.array([float(x) for x in line.strip().split('\t')[1:]])
+
     return means
 
 
@@ -1536,9 +1538,9 @@ def predictF1F2(phone, selectedpoles, selectedbandwidths, means, covs):
                     i = 0
                     j = 1
                     ## vector with current pole combination and associated bandwidths
-                    x = robjects.FloatVector([poles[i], poles[j], math.log(bandwidths[i]), math.log(bandwidths[j])])
+                    x = np.array([poles[i], poles[j], math.log(bandwidths[i]), math.log(bandwidths[j])])
                     ## calculate Mahalanobis distance between x and ANAE mean
-                    dist = robjects.r['mahalanobis'](x, means[vowel], covs[vowel])[0]
+                    dist = mahalanobis(x, means[vowel], linalg.inv(covs[vowel]))
                     ## append poles and bandwidths to list of values
                     ## (if F3 and bandwidth measurements exist, add to list of appended values)
                     if len(poles) > 2:

FAVE-extract/bin/remeasure.py

@@ -1,7 +1,8 @@
 #import ast
 import math
-import rpy2.robjects as robjects
-import rpy2.rinterface as rinterface
+import numpy as np
+from scipy import linalg
+from scipy.spatial.distance import mahalanobis as mahalanobis
 import sys
 import string
 
@@ -119,9 +120,9 @@ def pruneVowels(vowels, vowel, vowelMeans, vowelCovs, outlie):
     while not enough:
         outtokens = [ ]
         for token in vowels[vowel]:
-            x = robjects.FloatVector(token)
-            dist = robjects.r['mahalanobis'](x, vowelMeans[vowel], vowelCovs[vowel])[0]
-            if dist <= outlie:
+            x = np.array(token)
+            dist = mahalanobis(x, vowelMeans[vowel], linalg.inv(vowelCovs[vowel]))
+            if dist**2 <= outlie:
                 outtokens.append(token)
         if len(outtokens) >= 10:
             enough = True
@@ -144,19 +145,15 @@ def calculateVowelMeans(vowels):
     vowelMeans = {}
     vowelCovs = {}
     for vowel in vowels:
-        vF1 = robjects.FloatVector([F1 for [F1,F2,B1,B2,Dur] in vowels[vowel]])
-        vF2 = robjects.FloatVector([F2 for [F1,F2,B1,B2,Dur] in vowels[vowel]])
-        vB1 = robjects.FloatVector([B1 for [F1,F2,B1,B2,Dur] in vowels[vowel]])
-        vB2 = robjects.FloatVector([B2 for [F1,F2,B1,B2,Dur] in vowels[vowel]])
-        vDur = robjects.FloatVector([Dur for [F1,F2,B1,B2,Dur] in vowels[vowel]])
+        vF1 = np.array([F1 for [F1,F2,B1,B2,Dur] in vowels[vowel]])
+        vF2 = np.array([F2 for [F1,F2,B1,B2,Dur] in vowels[vowel]])
+        vB1 = np.array([B1 for [F1,F2,B1,B2,Dur] in vowels[vowel]])
+        vB2 = np.array([B2 for [F1,F2,B1,B2,Dur] in vowels[vowel]])
+        vDur = np.array([Dur for [F1,F2,B1,B2,Dur] in vowels[vowel]])
 
-        rcolMeans = robjects.r["colMeans"]
-        rcov = robjects.r["cov"]
-
-        measureMatrix = robjects.r["matrix"](vF1 + vF2 + vB1 + vB2 + vDur, ncol = 5)
 
-        vowelMeans[vowel] = rcolMeans(measureMatrix)
-        vowelCovs[vowel] = rcov(measureMatrix)
+        vowelMeans[vowel] = np.array([vF1.mean(), vF2.mean(), vB1.mean(), vB2.mean(), vDur.mean()])
+        vowelCovs[vowel] = np.cov(np.vstack((vF1, vF2, vB1, vB2, vDur)))
     #sys.stderr.write("Vowel means calculated\n")
     return vowelMeans, vowelCovs
 
@@ -203,7 +200,7 @@ def repredictF1F2(measurements, vowelMeans, vowelCovs, vowels):
                 values = [F1, F2, B1, B2, lDur]
                 outvalues = [F1, F2, F3, B1, B2, B3, lDur]
 
-                x = robjects.FloatVector(values)
+                x = np.array(values)
 
                 ##If there is only one member of a vowel category,
                 ##the covariance matrix will be filled with NAs
@@ -212,21 +209,21 @@ def repredictF1F2(measurements, vowelMeans, vowelCovs, vowels):
                     ## if no re-measurement is to take place for one of the three reasons below, the list of candidate measurements and nFormants
                     ## will be filled with four identical copies of the original measurement, all with a distance of zero
                     ## so that the original measurement is guaranteed to be re-selected
-                    if vowelCovs[vowel][0] is rinterface.NA_Real:
+                    if np.isnan(vowelCovs[vowel][0,0]):
                         valuesList.append([float(vm.f1), float(vm.f2), vm.f3, math.log(float(vm.b1)), math.log(float(vm.b2)), vm.b3, lDur])
                         distanceList.append(0)
                         nFormantsList.append(vm.nFormants)
                     elif len(vowels[vowel]) < 7:
                         valuesList.append([float(vm.f1), float(vm.f2), vm.f3, math.log(float(vm.b1)), math.log(float(vm.b2)), vm.b3, lDur])
                         distanceList.append(0)
                         nFormantsList.append(vm.nFormants)
-                    elif robjects.r['det'](vowelCovs[vowel])[0] == 0:  ## determinant of the covariance matrix is zero
+                    elif linalg.det(vowelCovs[vowel]) == 0:  ## determinant of the covariance matrix is zero
                         valuesList.append([float(vm.f1), float(vm.f2), vm.f3, math.log(float(vm.b1)), math.log(float(vm.b2)), vm.b3, lDur])
                         distanceList.append(0)
                         nFormantsList.append(vm.nFormants)
                     ## "real" re-measurement
                     else:
-                        dist = robjects.r['mahalanobis' ](x, vowelMeans[vowel], vowelCovs[vowel])[0]   
+                        dist = mahalanobis(x, vowelMeans[vowel], linalg.inv(vowelCovs[vowel]))   
                         valuesList.append(outvalues)
                         distanceList.append(dist)
                         nFormantsList.append(i + 3)  ## these are the formant setting used, not the actual number of formants returned