updated smma with pandas

Author: vincentropy

pyti/smoothed_moving_average.py

@@ -4,6 +4,7 @@
 from pyti.function_helper import fill_for_noncomputable_vals
 from six.moves import range
 
+import pandas as pd
 
 def smoothed_moving_average(data, period):
     """
@@ -13,8 +14,9 @@ def smoothed_moving_average(data, period):
     smma = avg(data(n)) - avg(data(n)/n) + data(t)/n
     """
     catch_errors.check_for_period_error(data, period)
-    smma = [((np.mean(data[idx-(period-1):idx+1]) -
-         (np.mean(data[idx-(period-1):idx+1])/period) +
-         data[idx])/period) for idx in range(0, len(data))]
-    smma = fill_for_noncomputable_vals(data, smma)
-    return smma
+    # smma = [((np.mean(data[idx-(period-1):idx+1]) -
+    #      (np.mean(data[idx-(period-1):idx+1])/period) +
+    #      data[idx])/period) for idx in range(0, len(data))]
+    # smma = fill_for_noncomputable_vals(data, smma)
+    series = pd.Series(data)
+    return series.ewm(alpha = 1.0/period).mean().values.flatten()

requirements.txt

@@ -1 +1,2 @@
 numpy
+pandas

setup.py

@@ -33,6 +33,6 @@
 
     packages=find_packages(exclude=['tests']),
 
-    install_requires=['numpy'],
+    install_requires=['numpy', 'pandas'],
 
 )