Time Series Plots

Python offers many custom options for plotting time series models. Both matplotlib and seaborn have time series plotting methods. Let's implement a time series plot using some stock data from quandl

import quandl
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt

%matplotlib inline
%config InlineBackend.figure_format = 'retina'

We use quandl to return stock data. The result is OHLCV metrics for both aggregate and adjusted at a daily level for all business days.

fb_stock = quandl.get('WIKI/FB', start_date='2017-01-01', end_date='2017-05-30', collapse='daily')
fb_stock.head()

	Open	High	Low	Close	Volume	Ex-Dividend	Split Ratio	Adj. Open	Adj. High	Adj. Low	Adj. Close	Adj. Volume
Date
2017-01-03	116.03	117.84	115.5100	116.86	20663912.0	0.0	1.0	116.03	117.84	115.5100	116.86	20663912.0
2017-01-04	117.55	119.66	117.2900	118.69	19630932.0	0.0	1.0	117.55	119.66	117.2900	118.69	19630932.0
2017-01-05	118.86	120.95	118.3209	120.67	19492150.0	0.0	1.0	118.86	120.95	118.3209	120.67	19492150.0
2017-01-06	120.98	123.88	120.0300	123.41	28545263.0	0.0	1.0	120.98	123.88	120.0300	123.41	28545263.0
2017-01-09	123.55	125.43	123.0400	124.90	22880360.0	0.0	1.0	123.55	125.43	123.0400	124.90	22880360.0

The best way to plot a time series chart with matplotlib is to leverage the date as the index and pass a dataframe with columns of interest to the plot method. The plot method will render all columns by color in the time series plot and provide a legennd for mapping. See example below.

fb_stock[['Open', 'High', 'Low' ,'Close']].plot(linestyle='-', linewidth=1.5, figsize=(10,6))
plt.xlabel('Daily Dates')
plt.ylabel('Stock Price')
plt.title('Facebook Daily Stock Price: Jan - May 2017')