Initial commit.

Author: matham

.gitignore

@@ -0,0 +1,6 @@
+*.pyc
+*.ini
+.project
+.pydevproject
+glitter/ffpyplayer
+glitter/libs

README.md

@@ -0,0 +1,3 @@
+CPL Lab, Cornell University
+Video tracking, scoring, and analysis software.
+LGPL3

glitter/__init__.py

@@ -0,0 +1,3 @@
+
+__version__ = 2.0
+__description__ = 'CPL Video tracking/scoring and analysis software.'

glitter/exporter.py

@@ -0,0 +1,151 @@
+import operator
+import sys
+if getattr(sys, 'frozen', False):
+    from collider import Collide2DPoly
+else:
+    from kivy.garden.collider import Collide2DPoly
+import itertools
+
+__all__ = ('DataList', )
+
+class DataList(list):
+
+    def __init__(self, d, score_type, name, func='', previous=None, *largs, **kwargs):
+        super(DataList, self).__init__(*largs, **kwargs)
+        self.score_type = score_type
+        ''' Possible score_type are xyt, xy, t, pts '''
+        self.previous = previous
+        self.name = name
+        self.func = func
+        self.pts = []
+        self.d = d
+    
+    def compare_constant(self, value, op):
+        if value.__class__ is not float and value.__class__ is not int:
+            raise Exception('Invalid data type, %s, given to %s comparison operator.' %(type(value), op))
+        if self.score_type != 'pts':
+            raise Exception("You can use the %s comparison operator only on pts." %op)
+        result = DataList(self.d, 't', self.name, '%s_%f_%s' % (op, float(value), self.name))
+        result.pts = self.pts
+        op = getattr(operator, op)
+        result[:] = [op(t, value) for t in self.d['_pts']]
+        return result
+    def __lt__(self, val):
+        return self.compare_constant(val, 'lt')
+    def __le__(self, val):
+        return self.compare_constant(val, 'le')
+    def __gt__(self, val):
+        return self.compare_constant(val, 'gt')
+    def __ge__(self, val):
+        return self.compare_constant(val, 'ge')
+    
+    def within_dist(self, value):
+        if value.__class__ is not float and value.__class__ is not int:
+            raise Exception('Invalid data type, %s, used in within_dist.' %type(value))
+        if self.score_type != 't':
+            raise Exception("You can use the within_dist operator only on t score types.")
+        result = DataList(self.d, 't', self.name, 'within_dist_%f_%s' % (float(value), self.name), self)
+        result.pts = self.pts
+        pts = self.d['_pts']
+        result[:] = [False, ] * len(self)
+        points = [pts[i] for i in range(len(self)) if self[i]]
+        for i in range(len(result)):
+            if value >= 0:
+                result[i] = any([point <= pts[i] < point + value for point in points])
+            else:
+                result[i] = any([point + value < pts[i] <= point for point in points])
+        return result
+        
+    def __or__(self, other):
+        if other.__class__ is not DataList:
+            raise Exception('Invalid data type given to or (|).')
+        score_types = (self.score_type, other.score_type)
+        if 'pts' in score_types:
+            raise Exception("You cannot 'or' pts type data.")
+        if score_types[0] == score_types[1]:
+            if score_types[0] != 't':
+                raise Exception("You can only 'or' identical data types if they are both of type t.")
+            if len(self) != len(other):
+                raise Exception("You can only 'or' data types t if they are both of the same length.")
+            result = DataList(self.d, self.score_type, self.name, 'or_'+other.name, self)
+            result.pts = self.pts
+            result.extend([self[i] or other[i] for i in range(len(self))])
+            return result
+        raise Exception("You cannot 'or' a "+score_types[0]+' type with a '+score_types[1]+' type.')
+    
+    def __invert__(self):
+        if self.score_type != 't':
+            raise Exception("You can only 'invert' t type data.")
+        result = DataList(self.d, self.score_type, self.name, 'invert_' + self.name, self)
+        result.pts = self.pts
+        result[:] = [not val for val in self]
+        return result
+        
+    def __and__(self, other):
+        if other.__class__ is not DataList:
+            raise Exception('Invalid data type given to and (&).')
+        score_types = (self.score_type, other.score_type)
+        if 'pts' in score_types:
+            raise Exception("You cannot 'and' pts type data.")
+        if score_types[0] == score_types[1]:
+            if score_types[0] != 't':
+                raise Exception("You can only 'and' identical data types if they are both of type t.")
+            if len(self) != len(other):
+                raise Exception("You can only 'and' data types t if they are both of the same length.")
+            result = DataList(self.d, self.score_type, self.name, 'and_'+other.name, self)
+            result.pts = self.pts
+            result.extend([self[i] and other[i] for i in range(len(self))])
+            return result
+        if 't' in score_types:
+            raise Exception("You cannot 'and' a "+score_types[0]+' type with a '+score_types[1]+' type.')
+        a, b = self, other
+        if score_types[0] == 'xyt':
+            a, b = b, a
+        collider = Collide2DPoly([p for p in itertools.chain.from_iterable(a)], cache=True)
+        result = DataList(b.d, 't', b.name, 'and_'+a.name, b)
+        result.pts = b.pts
+        result.extend([b[i] in collider for i in range(len(b))])
+        return result
+
+# A scaler result from an operation on array
+class DataResult(float):
+
+    def __new__(cls, previous=None, *largs, **kwargs):
+        obj = float.__new__(cls, *largs, **kwargs)
+        obj.previous = previous
+        return obj
+
+def esum(data):
+    return DataResult(data, sum(data))
+def emin(data):
+    return DataResult(data, min(data))
+def emax(data):
+    return DataResult(data, max(data))
+def elen(data):
+    return DataResult(data, len(data))
+
+def event(data):
+    if data.__class__ is not DataList:
+        raise Exception('Invalid data type given to event().')
+    if data.score_type != 't':
+        raise Exception('Only t type channels can be given to event().')
+    pts = data.d['_pts']
+    result = DataList(data.d, data.score_type, data.name, 'event', data)
+    start = -1
+    for i in range(len(data)):
+        if data[i] and start == -1:
+            start = i
+        if start != -1 and (i + 1 == len(data) or not data[i + 1]):
+            result.append(pts[i] - pts[start])
+            result.pts.append(start)
+            start = -1
+    return result
+
+def export_data(d, channels):
+    result = [0, ] * len(channels)
+    for i in range(len(channels)):
+        lines = channels[i].split('\n')
+        lines = [line for line in lines if line]
+        exec('\n'.join(lines[:-1]))
+        result[i] = eval(lines[-1])
+    return result

glitter/help.py

@@ -0,0 +1,36 @@
+
+
+
+
+__all__ = ('HelpDoc')
+
+
+HelpDoc = '''
+[b]Browsing bar:[/b]
+
+[b]Load video:[/b] Loads a video file. If set to load a default data file, it \
+also loads or create a data file with the same name as the video file. It closes\
+any previously opened video and data files. Current channels will not be deleted.
+ 
+[b]Load data file:[/b] Loads or creates a data file. All the recorded data\
+is stored in the PyTables HDF5 format data file.
+
+
+[b]Exporter:[/b]
+The following file level variables are available in the exporter:
+_version (provides the version of the program which wrote the file).
+_filename (the filename of the video file associated with the data file).
+_vid_info (a dict with information about the video file, e.g. frame size...).
+_user (the user that created the file).
+_ID (the filename of the video file associated with the data file).
+_complete (whether all the frames in the video has been watched).
+_pts (the list of the timestamps of the video).
+
+In addition, each channel's name defined in the data file is also available in \
+two flavors. For example, say you add a two channels, Head, and Tail in that \
+order. Then you can refer to them using either Head, Tail respectively, \
+or Head_0, Tail_1, respectively. That is, you can refer to each channel by \
+its name, or by appending _n, where n is the order number of the channel, \
+defined by the order in which the channels were created. This should resolve \
+naming conflicts.
+'''