From ab076f8c76189d12c6891844ffd6fa3e28d25d31 Mon Sep 17 00:00:00 2001
From: cerlymarco <36955807+cerlymarco@users.noreply.github.com>
Date: Thu, 3 Sep 2020 10:29:52 +0200
Subject: [PATCH] Release

---
 README.md               | 16 ++++++++--------
 setup.py                |  2 +-
 tsmoothie/smoother.py   |  2 +-
 tsmoothie/utils_func.py |  6 +++---
 4 files changed, 13 insertions(+), 13 deletions(-)

diff --git a/README.md b/README.md
index 5f7a582..efcb764 100644
--- a/README.md
+++ b/README.md
@@ -1,12 +1,12 @@
 # tsmoothie
 
-A python library for timeseries smoothing and outlier detection in a vectorized way.
+A python library for time-series smoothing and outlier detection in a vectorized way.
 
 ## Overview
 
-tsmoothie computes, in a fast and efficient way, the smoothing of single or multiple timeseries. 
+tsmoothie computes, in a fast and efficient way, the smoothing of single or multiple time-series. 
 
-The smoothing tecniques available are:
+The smoothing techniques available are:
 
 - Exponential Smoothing
 - Convolutional Smoothing with various window types (constant, hanning, hamming, bartlett, blackman)
@@ -17,7 +17,7 @@ The smoothing tecniques available are:
 - LOWESS 
 - Kalman Smoothing with customizable components (level, trend, seasonality, long seasonality) 
 
-tsmoothie provides the calculation of intervals as result of the smoothing process. This can be useful to identify outliers and anomalies in timeseries.
+tsmoothie provides the calculation of intervals as result of the smoothing process. This can be useful to identify outliers and anomalies in time-series.
 
 The interval types available are:
 
@@ -28,14 +28,14 @@ The interval types available are:
 
 The adoption of this type of intervals depends on the smoothing method used.
 
-tsmoothie can also carry out a sliding smoothing approach. This is possible splitting the timeseries into equal sized pieces and smoothing them independently. As always, this functionality is implemented in a vectorized way through the WindowWrapper class.
+tsmoothie can also carry out a sliding smoothing approach. This is possible splitting the time-series into equal sized pieces and smoothing them independently. As always, this functionality is implemented in a vectorized way through the WindowWrapper class.
 
 ## Media
 
 Blog Posts:
 
-- Timeseries Smoothing for better clustering (cooming soon)
-- Timeseries Smoothing for better forecasting (cooming soon)
+- [Time Series Smoothing for better Clustering](https://towardsdatascience.com/time-series-smoothing-for-better-clustering-121b98f308e8)
+- Time Series Smoothing for better Forecasting (coming soon)
 
 ## Installation
 
@@ -119,5 +119,5 @@ for i in range(3):
 
 ## References
 
-- Polynomial, Spline, Gaussian and Binner smoothing are carried out building a regression on custom basis expansions. These implementations are based on the amazing intutions of Matthew Drury available [here](https://github.com/madrury/basis-expansions/blob/master/examples/comparison-of-smoothing-methods.ipynb)
+- Polynomial, Spline, Gaussian and Binner smoothing are carried out building a regression on custom basis expansions. These implementations are based on the amazing intuitions of Matthew Drury available [here](https://github.com/madrury/basis-expansions/blob/master/examples/comparison-of-smoothing-methods.ipynb)
 - Time Series Modelling with Unobserved Components, Matteo M. Pelagatti
\ No newline at end of file
diff --git a/setup.py b/setup.py
index 0c2b637..e3ad171 100644
--- a/setup.py
+++ b/setup.py
@@ -3,7 +3,7 @@
 
 HERE = pathlib.Path(__file__).parent
 
-VERSION = '0.1.5'
+VERSION = '0.1.6'
 PACKAGE_NAME = 'tsmoothie'
 AUTHOR = 'Marco Cerliani'
 AUTHOR_EMAIL = 'cerlymarco@gmail.com'
diff --git a/tsmoothie/smoother.py b/tsmoothie/smoother.py
index 284810e..2f17e6d 100644
--- a/tsmoothie/smoother.py
+++ b/tsmoothie/smoother.py
@@ -278,7 +278,7 @@ def smooth(self, data):
         if self.window_type == 'ones':
             w = np.ones(self.window_len)
         else:
-            w = eval('np.'+self.window_type+'(window_len)')
+            w = eval('np.'+self.window_type+'(self.window_len)')
 
         if data.ndim == 2:
             pad_data = np.pad(data, ((self.window_len,self.window_len),(0,0)), mode='symmetric')
diff --git a/tsmoothie/utils_func.py b/tsmoothie/utils_func.py
index 80960d1..c3dc975 100644
--- a/tsmoothie/utils_func.py
+++ b/tsmoothie/utils_func.py
@@ -151,8 +151,8 @@ def create_windows(data, window_shape, step = 1, start_id = None, end_id = None)
     
     data = np.asarray(data)
 
-    if data.ndim == 0:
-        raise ValueError("Pass an object with more than one timesteps")
+    if data.ndim != 2:
+        raise ValueError("Pass a 2D array-like in the format (timestemps, series)")
 
     if window_shape < 1:
         raise ValueError("window_shape must be >= 1")
@@ -196,7 +196,7 @@ def sigma_interval(true, prediction, n_sigma):
         Upper bands.
     """
     
-    std = (true - prediction).std(1, keepdims=True)
+    std = np.nanstd(true - prediction, axis=1, keepdims=True)
     
     low = prediction - n_sigma*std
     up = prediction + n_sigma*std