Support for compacting binary data

1 files changed, 29 insertions, 0 deletions

diff --git a/binary_data.py b/binary_data.py
new file mode 100644
index 0000000..e5b47aa
--- /dev/null
+++ b/binary_data.py
@@ -0,0 +1,29 @@
+import numpy
+
+from util import load_binary_file
+from stats import iqr_remove_outliers, iqr_cutoff
+
+def compact_file(fname, scale=None, extend=1.5):
+    data = load_binary_file(fname)
+
+    if not scale is None:
+        data *= scale
+
+    data.sort()
+
+    iqr_min, iqr_max = iqr_cutoff(data, extend)
+
+    min_idx, max_idx = numpy.searchsorted(data, [iqr_min, iqr_max])
+
+    samples = data[min_idx:max_idx]
+
+    filtered = len(data) - len(samples)
+    max = samples[-1]
+    min = samples[0]
+    med = numpy.median(samples)
+    avg = numpy.mean(samples)
+
+    std = numpy.std(samples)
+    var = numpy.var(samples)
+
+    return [len(samples), filtered, max, avg, min, med, std, var]