NgramIterator.py

Author: Bill Majoros

BedReader.py

@@ -16,8 +16,9 @@
 #   fh : file handle
 # Instance Methods:
 #   reader=BedReader(filename)
-#   reader.close()
 #   record=reader.nextRecord() # Bed3Record or Bed6Record
+#   reader.close()
+#   list=BedReader.readAll(filename)
 # Class Methods:
 #   
 #=========================================================================
@@ -26,6 +27,17 @@ class BedReader:
     def __init__(self,filename):
         self.fh=open(filename,"r")
 
+    @classmethod
+    def readAll(cls,filename):
+        reader=BedReader(filename)
+        array=[]
+        while(True):
+            record=reader.nextRecord()
+            if(not record): break
+            array.append(record)
+        reader.close()
+        return array
+
     def close(self):
         self.fh.close()
 

FastaReader.py

@@ -25,7 +25,7 @@
 #   size=FastaReader.getSize(filename)
 #   FastaReader.readAll(filename) # returns hash : id->sequence
 #   FastaReader.readAllAndKeepDefs(filename) # returns hash : id->[def,seq]
-#   seq=FastaReader.firstSequence(filename)
+#   (defline,seq)=FastaReader.firstSequence(filename)
 #   [id,attribute_hash]=FastaReader.parseDefline(defline)
 #=========================================================================
 class FastaReader:

FastbTrack.py

@@ -22,7 +22,7 @@
 #   data : string (for discrete) or array of float (for continuous)
 #   deflineExtra : extra info for defline
 # Methods:
-#   track=FastbTrack(type,id,data,deflineExtra); # $type="discrete" or "continuous"
+#   track=FastbTrack(type,id,data,deflineExtra); # type="discrete" or "continuous"
 #   type=track.getType()
 #   data=track.getData()
 #   track.setSequence(string) # discrete
@@ -41,6 +41,7 @@
 #      "value" attribute added
 #   newTrack=track.slice(begin,end) # [begin,end) => end not inclusive
 #   meanValue=track.getMean(interval=None) # only for continuous data
+#   (maxValue,maxPos)=track.getMax(interval=None) # only for continuous data
 ######################################################################
 
 class FastbTrack:
@@ -177,6 +178,22 @@ def getMean(self,interval=None):
         mean=sum/L
         return mean
 
+    def getMax(self,interval=None):
+        data=self.data
+        L=len(data)
+        begin=0
+        end=L
+        if(interval is not None): 
+            begin=interval.begin
+            end=interval.end
+            L=end-begin
+        max=data[begin]; maxPos=begin
+        for i in range(begin+1,end):
+            if(data[i]>max):
+                max=data[i]
+                maxPos=i
+        return (max,maxPos)
+
     def anyZeroValues(self):
         data=self.data
         for x in data:

NgramIterator.py

@@ -0,0 +1,90 @@
+#!/usr/bin/env python
+#=========================================================================
+# This is OPEN SOURCE SOFTWARE governed by the Gnu General Public
+# License (GPL) version 3, as described at www.opensource.org.
+# Copyright (C)2017 William H. Majoros (martiandna@gmail.com).
+#=========================================================================
+from __future__ import (absolute_import, division, print_function, 
+   unicode_literals, generators, nested_scopes, with_statement)
+from builtins import (bytes, dict, int, list, object, range, str, ascii,
+   chr, hex, input, next, oct, open, pow, round, super, filter, map, zip)
+# The above imports should allow this program to run in both Python 2 and
+# Python 3.  You might need to update your version of module "future".
+
+######################################################################
+#
+# NgramIterator.py bmajoros
+#
+# Attributes:
+#   array ngram : array of integer indices into alphabet string
+#   string alphabet
+# Methods:
+#   ngramIterator=NgramIterator("ATCG",N)
+#   string=ngramIterator.nextString() # returns None if no more
+#   ngramIterator.reset()
+# Private methods:
+#   self.ngramToString()
+######################################################################
+
+class NgramIterator:
+#---------------------------------------------------------------------
+#                           PUBLIC METHODS
+#---------------------------------------------------------------------
+# ngramIterator=NgramIterator("ATCG",N)
+    def __init__(self,alphabet,N):
+        ngram=[]
+        alphaSize=len(alphabet)
+        for i in range(N-1): ngram.append(0)
+        if(N>0): ngram.append(-1)
+        self.alphabet=alphabet
+        self.ngram=ngram
+#---------------------------------------------------------------------
+# ngramIterator.reset()
+    def reset(self):
+        ngram=self.ngram
+        L=len(ngram)
+        for i in range(L-1): ngram[i]=0
+        ngram[L-1]=-1
+#---------------------------------------------------------------------
+# string=ngramIterator.nextString() # returns undef if no more
+    def nextString(self):
+        alphabet=self.alphabet
+        ngram=self.ngram
+        if(ngram is None): return None
+        L=len(ngram)
+        if(L==0):
+            self.ngram=None
+            return ""
+        alphaSize=len(alphabet)
+        for i in range(L-1,-1,-1): # for($i=$len-1 ; $i>=0 ; --$i)
+            ngram[i]+=1
+            index=ngram[i]
+            if(index<alphaSize): return self.ngramToString()
+            ngram[i]=0
+        return None
+#---------------------------------------------------------------------
+#---------------------------------------------------------------------
+#---------------------------------------------------------------------
+#---------------------------------------------------------------------
+#---------------------------------------------------------------------
+
+
+
+
+
+
+#---------------------------------------------------------------------
+#                         PRIVATE METHODS
+#---------------------------------------------------------------------
+# self.ngramToString()
+    def ngramToString(self):
+        ngram=self.ngram
+        alphabet=self.alphabet
+        length=len(ngram)
+        string=""
+        for i in range(length):
+            index=ngram[i]
+            string+=alphabet[index:index+1]
+        return string
+
+

template.py

@@ -10,6 +10,12 @@
    chr, hex, input, next, oct, open, pow, round, super, filter, map, zip)
 # The above imports should allow this program to run in both Python 2 and
 # Python 3.  You might need to update your version of module "future".
+import sys
+import ProgramName
+
+if(len(sys.argv)!=2):
+    exit(ProgramName.get()+" <>\n")
+()=sys.argv[1:]