# Smoothing in Python

Learn how to perform smoothing using various methods in Python.

If you're using Dash Enterprise's Data Science Workspaces, you can copy/paste any of these cells into a
Workspace Jupyter notebook.

Alternatively, download this entire tutorial as a Jupyter notebook and import it into your Workspace.

Find out if your company is using Dash Enterprise.

## New to Plotly?

Plotly is a free and open-source graphing library for Python. We recommend you read our Getting Started guide for the latest installation or upgrade instructions, then move on to our Plotly Fundamentals tutorials or dive straight in to some Basic Charts tutorials.

```
import plotly.graph_objects as go
import numpy as np
import pandas as pd
import scipy
from scipy import signal
```

#### Savitzky-Golay Filter¶

`Smoothing`

is a technique that is used to eliminate noise from a dataset. There are many algorithms and methods to accomplish this but all have the same general purpose of 'roughing out the edges' or 'smoothing' some data.

There is reason to smooth data if there is little to no small-scale structure in the data. The danger to this thinking is that one may skew the representation of the data enough to change its percieved meaning, so for the sake of scientific honesty it is an imperative to at the very minimum explain one's reason's for using a smoothing algorithm to their dataset.

In this example we use the Savitzky-Golay Filter, which fits subsequents windows of adjacent data with a low-order polynomial.

```
import plotly.graph_objects as go
import numpy as np
import pandas as pd
import scipy
from scipy import signal
np.random.seed(1)
x = np.linspace(0, 10, 100)
y = np.sin(x)
noise = 2 * np.random.random(len(x)) - 1 # uniformly distributed between -1 and 1
y_noise = y + noise
fig = go.Figure()
fig.add_trace(go.Scatter(
x=x,
y=y,
mode='markers',
marker=dict(size=2, color='black'),
name='Sine'
))
fig.add_trace(go.Scatter(
x=x,
y=y_noise,
mode='markers',
marker=dict(
size=6,
color='royalblue',
symbol='circle-open'
),
name='Noisy Sine'
))
fig.add_trace(go.Scatter(
x=x,
y=signal.savgol_filter(y,
53, # window size used for filtering
3), # order of fitted polynomial
mode='markers',
marker=dict(
size=6,
color='mediumpurple',
symbol='triangle-up'
),
name='Savitzky-Golay'
))
fig.show()
```

#### Triangular Moving Average¶

Another method for smoothing is a moving average. There are various forms of this, but the idea is to take a window of points in your dataset, compute an average of the points, then shift the window over by one point and repeat. This will generate a bunch of points which will result in the `smoothed`

data.

Let us look at the common `Simple Moving Average`

first. In the 1D case we have a data set of $N$ points with y-values $y_1, y_2, ..., y_N$. Setting our window size to $n < N$, the new $i^{th}$ y-value after smoothing is computed as:

In the `Triangular Moving Average`

, two simple moving averages are computed on top of each other, in order to give more weight to closer (adjacent) points. This means that our $SMA_i$ are computed then a Triangular Moving Average $TMA_i$ is computed as:

```
def smoothTriangle(data, degree):
triangle=np.concatenate((np.arange(degree + 1), np.arange(degree)[::-1])) # up then down
smoothed=[]
for i in range(degree, len(data) - degree * 2):
point=data[i:i + len(triangle)] * triangle
smoothed.append(np.sum(point)/np.sum(triangle))
# Handle boundaries
smoothed=[smoothed[0]]*int(degree + degree/2) + smoothed
while len(smoothed) < len(data):
smoothed.append(smoothed[-1])
return smoothed
fig = go.Figure()
fig.add_trace(go.Scatter(
x=x,
y=y,
mode='markers',
marker=dict(
size=2,
color='rgb(0, 0, 0)',
),
name='Sine'
))
fig.add_trace(go.Scatter(
x=x,
y=y_noise,
mode='markers',
marker=dict(
size=6,
color='#5E88FC',
symbol='circle-open'
),
name='Noisy Sine'
))
fig.add_trace(go.Scatter(
x=x,
y=smoothTriangle(y_noise, 10), # setting degree to 10
mode='markers',
marker=dict(
size=6,
color='#C190F0',
symbol='triangle-up'
),
name='Moving Triangle - Degree 10'
))
fig.show()
```

### What About Dash?¶

Dash is an open-source framework for building analytical applications, with no Javascript required, and it is tightly integrated with the Plotly graphing library.

Learn about how to install Dash at https://dash.plot.ly/installation.

Everywhere in this page that you see `fig.show()`

, you can display the same figure in a Dash application by passing it to the `figure`

argument of the `Graph`

component from the built-in `dash_core_components`

package like this:

```
import plotly.graph_objects as go # or plotly.express as px
fig = go.Figure() # or any Plotly Express function e.g. px.bar(...)
# fig.add_trace( ... )
# fig.update_layout( ... )
import dash
import dash_core_components as dcc
import dash_html_components as html
app = dash.Dash()
app.layout = html.Div([
dcc.Graph(figure=fig)
])
app.run_server(debug=True, use_reloader=False) # Turn off reloader if inside Jupyter
```