# HG changeset patch
# User kellrott
# Date 1353220960 18000
# Node ID 1f93906c2945c1a8f1829691f750ede80792fc75

Uploaded

diff -r 000000000000 -r 1f93906c2945 tabular_label_convert/._tabular_label_convert.py
Binary file tabular_label_convert/._tabular_label_convert.py has changed
diff -r 000000000000 -r 1f93906c2945 tabular_label_convert/tabular_label_convert.py
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/tabular_label_convert/tabular_label_convert.py	Sun Nov 18 01:42:40 2012 -0500
@@ -0,0 +1,209 @@
+#!/usr/bin/env python
+
+import os
+import csv
+import sys
+import array
+import math
+from copy import copy
+from argparse import ArgumentParser
+
+
+class FloatMatrix:
+    def __init__(self):
+        self.corner_name = "probe"
+        self.data = None
+        self.nrows = None
+        self.ncols = None
+        self.rowmap = None
+        self.colmap = None
+
+    def read(self, handle):
+        header = None
+        for line in handle:
+            row = line.rstrip().split("\t")
+            if header is None:
+                header = row
+                self.data = array.array("f")
+                self.colmap = {}
+                self.rowmap = {}
+                self.ncols = len(row) - 1
+                self.nrows = 0
+                for i, c in enumerate(row[1:]):
+                    self.colmap[c] = i
+            else:
+                if len(row) - 1 != self.ncols:
+                    raise DataException("Misformed matrix")
+                self.rowmap[row[0]] = len(self.rowmap)
+                a = []
+                for v in row[1:]:
+                    try:
+                        a.append(float(v))
+                    except ValueError:
+                        a.append(float('Nan'))
+                self.data.extend(a)
+                self.nrows += 1
+
+    def init_blank(self, rows, cols):
+        self.data = array.array("f")
+        self.colmap = {}
+        for i,c in enumerate(cols):
+            self.colmap[c] = i
+        self.rowmap = {}
+        for i,r in enumerate(rows):
+            self.rowmap[r] = i
+        self.ncols = len(cols)
+        self.nrows = len(rows)
+        for i in range(self.nrows):
+            self.data.extend([float('nan')] * self.ncols)
+
+    def get_value(self, row_name, col_name):
+        return self.data[ self.rowmap[row_name] * self.ncols + self.colmap[col_name] ]
+
+    def set_value(self, row_name, col_name, value):
+        self.data[ self.rowmap[row_name] * self.ncols + self.colmap[col_name] ] = value
+    
+    def get_row(self, row_name):
+        return self.data[ self.rowmap[row_name] * self.ncols :  (self.rowmap[row_name]+1) * self.ncols ]
+
+    def get_cols(self):
+        out = self.colmap.keys()
+        return sorted(out, key=self.colmap.get)
+    
+    def has_row(self, row):
+        return row in self.rowmap 
+
+    def has_col(self, col):
+        return col in self.colmap 
+
+    def get_rows(self):
+        out = self.rowmap.keys()
+        return sorted(out, key=self.rowmap.get)
+    
+    def write(self, handle, missing='NA'):
+        write = csv.writer(handle, delimiter="\t", lineterminator='\n')
+        col_list = self.get_cols()
+        
+        write.writerow([self.corner_name] + col_list)
+        for rowName in self.rowmap:
+            out = [rowName]
+            row = self.get_row(rowName)
+            for col in col_list:
+                val = row[self.colmap[col]]
+                if val is None or math.isnan(val):
+                    val = missing
+                else:
+                    val = "%.5f" % (val)
+                out.append(val)
+            write.writerow(out)        
+            
+
+def median(inList):
+    """calculates median"""
+    cList = copy(inList)
+    if len(cList) == 0:
+        median = float("nan")
+    elif len(cList) == 1:
+        return cList[0]
+    else:
+        cList.sort()
+        if len(cList)%2 == 1:
+            median = cList[len(cList)/2]
+        else:
+            median = (cList[len(cList)/2]+cList[(len(cList)/2)-1])/2.0
+    return (median)
+
+def mean(inList):
+    return sum(inList) / float(len(inList))
+
+def aliasRemap(inputMatrix, aliasMap, mode, combine_func):
+    """
+    Given a inputMatrix and an alias map, create a new genomic matrix 
+    with the probes from the original matrix remapped to the connected aliases
+    from the map
+    """
+    
+    if mode == "row":
+        i_am = {}
+        for label in aliasMap:
+            if inputMatrix.has_row(label):
+                for alias in aliasMap[label]:
+                    if alias not in i_am:
+                        i_am[alias] = {}
+                    i_am[alias][label] = True
+            
+        out = FloatMatrix()
+        out.init_blank( rows=i_am.keys(), cols=inputMatrix.get_cols() )
+        for a in i_am:
+            for sample in inputMatrix.get_cols():
+                o = []
+                for p in i_am[a]:
+                    if inputMatrix.has_row(p):
+                        o.append( inputMatrix.get_value( col_name=sample, row_name=p) )
+                if len(o):
+                    out.set_value(col_name=sample, row_name=a, value=combine_func(o))
+        return out
+
+    if mode == "col":
+        i_am = {}
+        for label in aliasMap:
+            if inputMatrix.has_col(label):
+                for alias in aliasMap[label]:
+                    if alias not in i_am:
+                        i_am[alias] = {}
+                    i_am[alias][label] = True
+            
+        out = FloatMatrix()
+        out.init_blank( cols=i_am.keys(), rows=inputMatrix.get_rows() )
+        for a in i_am:
+            for r in inputMatrix.get_rows():
+                o = []
+                for label in i_am[a]:
+                    if inputMatrix.has_col(label):
+                        o.append( inputMatrix.get_value( row_name=r, col_name=label) )
+                if len(o):
+                    out.set_value(col_name=a, row_name=r, value=combine_func(o))
+        return out
+
+
+combine_map = {
+    "mean" : mean,
+    "median" : median,
+    "max" : max,
+    "min" : min
+}
+
+if __name__ == "__main__":
+    parser = ArgumentParser()
+    
+    parser.add_argument("-m", "--mode", dest="mode", help="Row/Column mode", default="row")
+    parser.add_argument("-c", "--combine", dest="combine", help="Value Combine Method", default="mean")
+    parser.add_argument("-o", "--output", help="Output file", default=None)
+    parser.add_argument("inTab", help="Input tabular file", default=None)
+    parser.add_argument("aliasMap", help="Input alias map", default=None)
+    
+    args = parser.parse_args()
+
+    mtx = FloatMatrix()
+    handle = open(args.inTab)
+    mtx.read(handle)
+    handle.close()
+    
+    aliasMap = {}
+    handle = open(args.aliasMap)
+    for line in handle:
+        tmp = line.rstrip().split("\t")
+        if tmp[0] not in aliasMap:
+            aliasMap[tmp[0]] = {tmp[1] : True}
+        else:
+            aliasMap[tmp[0]][tmp[1]] = True
+    handle.close()
+    
+    out = aliasRemap(mtx, aliasMap, args.mode, combine_map[args.combine])
+    if args.output is None:
+        handle = sys.stdout
+    else:
+        handle = open(args.output, "w")
+    out.write(handle)
+    handle.close()
+    
diff -r 000000000000 -r 1f93906c2945 tabular_label_convert/tabular_label_convert.xml
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/tabular_label_convert/tabular_label_convert.xml	Sun Nov 18 01:42:40 2012 -0500
@@ -0,0 +1,65 @@
+<tool id="tabular_label_convert" name="Tabular Label Convert" version="1.0.0">
+  <description>Tabular Label Convert</description>
+  <command interpreter="python">tabular_label_convert.py -m ${dim} -c ${merge_mode} $intab $aliasMap -o $outfile</command>
+  <inputs>
+	  <param name="intab" type="data" format="tabular" label="Tabular Input"/>
+	  <param name="aliasMap" type="data" format="tabular" label="AliasMap"/>
+	  <param name="dim" type="select" label="Merge Across">
+	  	<option value="row">Rows</option>
+	  	<option value="col">Columns</option>
+	  </param>
+	  <param name="merge_mode" type="select" label="Merge Mode">
+	  	<option value="mean">Mean</option>
+	  	<option value="median">Median</option>
+	  	<option value="min">Min</option>
+	  	<option value="max">Max</option>
+	  </param>
+  </inputs>
+  <outputs>
+      <data name="outfile" format="tabular"/>
+  </outputs>
+  <help>
+This tool takes a tabular format file of numerical values and converts the labels of the 
+rows or columns using an alias map. If two or more labels map to the same alias, then the mean/median/min/max
+value is selected to represent the new alias.
+
+Input tabular format
+    A tab seperated matrix for floats, with the row and column serving as labels.
+
+
++-----------+----------------------+-----------------------+-----------------------+-----------------------+
+|probe      |  sample_1            |  sample_2             |   sample_3            |  sample_4             |
++-----------+----------------------+-----------------------+-----------------------+-----------------------+
+|207189_s_at|     -0.210051        | -0.492633             | -0.133921             |  -0.462803            |
++-----------+----------------------+-----------------------+-----------------------+-----------------------+
+|207190_at  |    -0.427553         |  -0.021174            |-0.398800              | -0.518978             |
++-----------+----------------------+-----------------------+-----------------------+-----------------------+
+|216013_at  |     -0.764110        | -0.911805             | -1.450937             |  -1.201317            |
++-----------+----------------------+-----------------------+-----------------------+-----------------------+
+
+Alias map format
+    A two column tab seperated files, the first column is the source label with the second column the destination alias.
+
++-------------+-----+
+|216013_at    |ZXDB |
++-------------+-----+
+|207189_s_at  |ZZEF1|
++-------------+-----+
+|207190_at    |ZZEF1|
++-------------+-----+
+
+Results
+    With the doing row relabel with 'mean' combination the matrix becomes:
+
+
++-----------+----------------------+-----------------------+-----------------------+-----------------------+
+|probe      |  sample_1            |  sample_2             |   sample_3            |  sample_4             |
++-----------+----------------------+-----------------------+-----------------------+-----------------------+
+|ZZEF1      |     -0.3188020       | -0.2569035            | -0.2663605            | -0.4908905            |
++-----------+----------------------+-----------------------+-----------------------+-----------------------+
+|ZXDB       |     -0.764110        | -0.911805             | -1.450937             |  -1.201317            |
++-----------+----------------------+-----------------------+-----------------------+-----------------------+
+
+
+  </help>
+</tool>