From 56267a16bab916d5baa4fa7344520e395110fd90 Mon Sep 17 00:00:00 2001
From: "Bjoern B. Brandenburg" <bbb@cs.unc.edu>
Date: Mon, 21 Feb 2011 17:20:59 -0500
Subject: add data filter based on standard deviation

---
 binary_data.py | 22 +++++++++++++++-------
 1 file changed, 15 insertions(+), 7 deletions(-)

(limited to 'binary_data.py')

diff --git a/binary_data.py b/binary_data.py
index 6108816..152afb7 100644
--- a/binary_data.py
+++ b/binary_data.py
@@ -4,7 +4,8 @@ from util import load_binary_file
 from stats import iqr_remove_outliers, iqr_cutoff
 
 
-def get_data(fname, scale, extent, cutoff=None, maxval=1000.0):
+def get_data(fname, scale, extent, cutoff=None, maxval=1000.0,
+             stdev=False):
     data = load_binary_file(fname)
 
     if cutoff and len(data) > cutoff:
@@ -15,15 +16,22 @@ def get_data(fname, scale, extent, cutoff=None, maxval=1000.0):
 
     data.sort()
 
-    if extent:
-        iqr_min, iqr_max = iqr_cutoff(data, extent)
+    if stdev:
+        # standard deviations filter
+        std = numpy.std(data, ddof=1)
+        avg = numpy.mean(data)
+        lower = avg - std * extent
+        upper = avg + std * extent
+    elif extent:
+        # IQR filter
+        lower, upper = iqr_cutoff(data, extent)
     else:
-        iqr_min = 0
-        iqr_max = maxval
+        lower = 0
+        upper = maxval
 
-    min_idx, max_idx = numpy.searchsorted(data, [iqr_min, iqr_max])
+    min_idx, max_idx = numpy.searchsorted(data, [lower, upper])
 
-    return [data, max_idx, min_idx, iqr_max, iqr_min]
+    return [data, max_idx, min_idx, upper, lower]
 
 
 def compact_file(*args, **kargs):
-- 
cgit v1.2.2