--- a
+++ b/Semantic Features/GetXws.m
@@ -0,0 +1,79 @@
+function [XwsRaw numSegments] = GetXws(instanceID)
+%GetXws Gets the X matrix of feature data from the weak segmentations
+%   Just reads in data from CSVs created by Patrick's extraction software
+
+%get features from our weak segmentors %FYI 4 Otsu and 6 Region Growing
+
+fprintf('Reading calculated feature data from text files.\n');
+featureDirectory = 'C:\Users\esmith2\Documents\Ethans Code\gitCode\Image features calculation code\Features for Production\2K image dataset\';
+%featureDirectory = 'C:\Users\esmith2\Documents\Ethans Code\gitCode\Image features calculation code\Features for Production\6K org image dataset\';
+%featureDirectory = 'C:\Ethan\repos\Image features calculation code\Features for Production\\6K org image dataset\';
+
+%get otsu features
+%[markovValues markovHeaders]
+fprintf('Reading Otsu segmentation features\n');
+fprintf('\tReading intensity values.\n');
+[intensityValuesO intensityHeadersO intensityDCMOrderO] = parseCSV(strcat(featureDirectory, 'otsu_nodules_intensity_features.txt'));
+fprintf('\tReading gabor values.\n');
+[gaborValuesO gaborHeadersO gaborDCMOrderO] = parseCSV(strcat(featureDirectory, 'otsu_nodules_gab_features.txt'));
+fprintf('\tReading haralick values.\n');
+[haralickValuesO haralickHeadersO haralickDCMOrderO] = parseCSV(strcat(featureDirectory, 'otsu_nodules_har_features.txt'));
+fprintf('\tReading shape values.\n');
+[shapeValuesO shapeHeadersO shapeDCMOrderO] = parseCSV(strcat(featureDirectory, 'otsu_nodules_shape_features.txt'));
+
+%get region grown features and horizontally concatenate them. 
+%[markovValues markovHeaders]
+fprintf('Reading Region Growing segmentation features\n');
+fprintf('\tReading intensity values.\n');
+[intensityValuesRG intensityHeadersRG intensityDCMOrderRG] = parseCSV(strcat(featureDirectory, 'rg_nodules_intensity_features.txt'));
+fprintf('\tReading gabor values.\n');
+[gaborValuesRG gaborHeadersRG gaborDCMOrderRG] = parseCSV(strcat(featureDirectory, 'rg_nodules_gab_features.txt'));
+fprintf('\tReading haralick values.\n');
+[haralickValuesRG haralickHeadersRG haralickDCMOrderRG] = parseCSV(strcat(featureDirectory, 'rg_nodules_har_features.txt'));
+fprintf('\tReading shape values.\n');
+[shapeValuesRG shapeHeadersRG shapeDCMOrderRG] = parseCSV(strcat(featureDirectory, 'rg_nodules_shape_features.txt'));
+
+%concatenate
+intensityValues = [intensityValuesO, intensityValuesRG];
+gaborValues = [gaborValuesO, gaborValuesRG];
+haralickValues = [haralickValuesO, haralickValuesRG];
+shapeValues = [shapeValuesO, shapeValuesRG];
+
+%This should happen before the reshaping of the matrices I think
+%Rearrange the calculated features so they match with the appropriate rating
+fprintf('Reorganizing data to match ratings.\n');
+newRowOrder = 0;
+for i = 1:size(instanceID)
+    row = find( instanceID(i) == intensityDCMOrderO);
+    if ~isempty(row)
+        newRowOrder = vertcat(newRowOrder, row); 
+    end
+end
+newRowOrder = newRowOrder(2:end,:);
+
+%apply reordering
+intensityValues     = intensityValues(newRowOrder,:);
+gaborValues         = gaborValues(newRowOrder,:);
+haralickValues      = haralickValues(newRowOrder,:);
+shapeValues         = shapeValues(newRowOrder,:);
+
+XwsRaw = [intensityValues, gaborValues, haralickValues, shapeValues]; 
+
+%get number of segments
+%Count Otsu segments
+i = 1;
+while intensityHeadersO{1}(end) == intensityHeadersO{i+1}(end)
+    i = i+1;
+end
+numSegments = length(intensityHeadersO)/i;
+%count region growing segments
+i = 1; 
+while intensityHeadersRG{1}(end) == intensityHeadersRG{i+1}(end)
+    i = i+1;
+end
+numSegments = numSegments + length(intensityHeadersRG)/i;
+
+%unused method of getting the largest boundary from each slice
+%largeInstances = intensityValues(:,end);
+
+end
\ No newline at end of file