improve performance and add new feature

Author: akialbz

dimspy/models/peak_matrix.py

@@ -401,6 +401,7 @@ def rsd(self, *args, **kwargs):
 
         :param args: tags or untyped tag values for RSD calculation, no value = calculate over all samples
         :param kwargs: typed tags for RSD calculation, , no value = calculate over all samples
+        :param on_attr: calculate RSD on given attribute. Default = "intensity"
         :param flagged_only: whether to calculate on flagged peaks only. Default = True
         :type: numpy array
 
@@ -413,6 +414,7 @@ def rsd(self, *args, **kwargs):
         corresponding rsd value will be set to np.nan.
 
         """
+        on_attr = kwargs.pop('on_attr') if kwargs.has_key('on_attr') else 'intensity'
         flagged_only = kwargs.pop('flagged_only') if kwargs.has_key('flagged_only') else True
 
         if self.shape[0] < 2:
@@ -423,7 +425,7 @@ def rsd(self, *args, **kwargs):
               unmask_peakmatrix(self, *args, **kwargs)) as m:
             if m.shape[0] == 0: raise AttributeError('peak matrix does not have label(s) [%s]' %
                                                      join(map(lambda x: str(x)[1:-1], (args, kwargs)), ', '))
-            ints = m.attr_matrix('intensity', flagged_only)
+            ints = m.attr_matrix(on_attr, flagged_only)
             rsd = m._present_std(ints, 0, flagged_only) / m._present_mean(ints, 0, flagged_only) * 100
 
         rsd[np.where(map(lambda x: len(set(x[np.nonzero(x)])) == 1, ints.T))] = np.nan  # only one valid value

dimspy/portals/mzml_portal.py

@@ -14,16 +14,21 @@
 import pymzml
 import numpy as np
 import zipfile
+from copy import deepcopy
 from dimspy.models.peaklist import PeakList
 from dimspy.experiment import mz_range_from_header
 
 
 class Mzml:
-    def __init__(self, filename="", archive=None):
+    def __init__(self, filename="", archive=None, preload=True):
         self.filename = filename
         self.archive = archive
+        self._preload = preload
+        self._cache = None
 
     def run(self):
+        if self._cache is not None: return self._cache
+
         if not self.filename.lower().endswith(".mzml") and not self.filename.lower().endswith(".mzml.gz") and not self.filename.lower().endswith(".zip"):
             raise IOError('Incorrect file format for mzML parser')
         if self.archive is not None:
@@ -32,11 +37,15 @@ def run(self):
             zf = zipfile.ZipFile(self.archive, 'r')
             if self.filename not in zf.namelist():
                 raise IOError("{} does not exist in zip file".format(self.filename))
-            return pymzml.run.Reader('', file_object=zf.open(self.filename))
+            dat = pymzml.run.Reader('', file_object=zf.open(self.filename))
+            if self._preload: dat = self._cache = tuple(map(deepcopy, dat))
+            return dat
         elif self.filename.lower().endswith(".mzml") or self.filename.lower().endswith(".mzml.gz"):
             if not os.path.isfile(self.filename):
                 raise IOError("{} does not exist".format(self.filename))
-            return pymzml.run.Reader(self.filename)
+            dat = pymzml.run.Reader(self.filename)
+            if self._preload: dat = self._cache = tuple(map(deepcopy, dat))
+            return dat
         else:
             return None
 

dimspy/process/peak_filters.py

@@ -92,21 +92,22 @@ def filter_mz_ranges(pl, mz_remove_rngs, flag_name='mz_range_remove_flag', flag_
 
 
 # PeakMatrix filters
-def filter_rsd(pm, rsd_threshold, qc_tag, flag_name='rsd_flag'):
+def filter_rsd(pm, rsd_threshold, qc_tag, on_attr = 'intensity', flag_name='rsd_flag'):
     """
     PeakMatrix RSD filter.
 
     :param pm: the target peak matrix
     :param rsd_threshold: threshold of the RSD of the QC samples
     :param qc_tag: tag (label) to unmask qc samples
+    :param on_attr: calculate RSD on given attribute. Default = "intensity"
     :param flag_name: name of the new flag. Default = 'rsd_flag'
     :rtype: PeakMatrix object
 
     This filter will calculate the RSD values of the QC samples. A peak with a QC RSD value larger than the
     threshold will be unflagged.
 
     """
-    rsd_values = pm.rsd(qc_tag)
+    rsd_values = pm.rsd(qc_tag, on_attr = on_attr)
     if np.any(np.isnan(rsd_values)):
         logging.warning('nan found in QC rsd values, filter might not work properly')