Yousef Time Series Analysis in Python 2020
Uploaded by
yousef shabanYousef Time Series Analysis in Python 2020
Uploaded by
yousef shabanhttps://www.kaggle.
com/jagangupta/time-series-basics-exploring-traditional-ts
https://www.kaggle.com/thebrownviking20/everything-you-can-do-with-a-time-series
02 Setting Up the Environment
Why jupyter
Problem: Install three packages if you don’t have them.
1. matplotlib – a plotting library for Python and its numerical mathematics
extension NumPy
2. seaborn – plotting library for Python and modeling in particular that is
based on matplotlib
3. arch - a statistical modelling package that we will employ throughout the
course
Solution:
1. Enter the Anaconda Prompt, as seen in the lecture and write:
pip install matplotlib
or
conda install matplotlib
You should receive a message that matplotlib is already installed (it comes
with Anaconda, remember?)
2. Enter the Anaconda Prompt, as seen in the lecture and write:
pip install seaborn
or
conda install seaborn
3. Enter the Anaconda Prompt, as seen in the lecture and write:
pip install arch
03 Introduction to Time Series in Python
013 Section-3-Introduction-to-Time-Series-Template
https://www.dropbox.com/s/59zz1ch8rabk5by/Section%203%20-%20Introduction%20to%20Time
%20Series%20-%20Template.ipynb?dl=0
013 Section-3-Introduction-to-Time-Series-Completed
https://www.dropbox.com/s/3ssi6wr8cuw1nkn/Section%203%20-%20Introduction%20to%20Time
%20Series%20-%20Completed%20.ipynb?dl=0
04 Creating a Time Series Object in Python
df_comp.dtypes
Out[8]:
date datetime64[ns]
spx float64
dax float64
ftse float64
nikkei float64
dtype: object
df_comp.date.describe()
Out[9]:
count 6269
unique 6269
top 1999-08-23 00:00:00
freq 1
first 1994-01-07 00:00:00
last 2018-01-29 00:00:00
Name: date, dtype: object
#Handling mssing values on DataSets
df_comp.isna()
df_comp.isna().sum()
df_comp.spx = df_comp.spx.fillna(method="ffill")
df_comp.isna().sum()
df_comp.ftse = df_comp.ftse.fillna(method="bfill")
df_comp.isna().sum()
df_comp.dax = df_comp.dax.fillna(value=df_comp.dax.mean())
df_comp.isna().sum()
017 Section-4-Creating-a-Time-Series-Object-in-Python-Template
https://www.dropbox.com/s/x50lb1h3dsb46bx/Section%204%20-%20Creating%20a%20Time
%20Series%20Object%20in%20Python%20-%20Template%20.ipynb?dl=0
017 Section-4-Creating-a-Time-Series-Object-in-Python-Completed
https://www.dropbox.com/s/nnrun96i0lvacs7/Section%204%20-%20Creating%20a%20Time
%20Series%20Object%20in%20Python%20-%20Completed%20.ipynb?dl=0
023 Appendix-Updating-the-Dataset
https://www.dropbox.com/s/b3eec9252hqqgkl/Updating%20The%20Data.ipynb?dl=0
05 Working with Time Series in Python
024 Section-5-Working-with-Time-Series-in-Python-Template
https://www.dropbox.com/s/qek5iy93vn9s058/Section%205%20-%20Working%20with%20Time
%20Series%20in%20Python%20-%20Template%20.ipynb?dl=0
024 Section-5-Working-with-Time-Series-in-Python-Completed
https://www.dropbox.com/s/yaz9ue79ao52wrv/Section%205%20-%20Working%20with%20Time
%20Series%20in%20Python%20-%20Completed%20.ipynb?dl=0
06 Picking the Correct Model
07 Modeling Autoregression The AR Model
08 Adjusting to Shocks The MA Model
046 Section-8-The-MA-Model-Template
https://www.dropbox.com/s/ikdi97raw63afmm/Sec
tion%208%20-%20The%20MA%20Model%20-
%20Template.ipynb?dl=0
046 Section-8-The-MA-Model-Completed
https://www.dropbox.com/s/jszpd43gay51jys/Sectio
n%208%20-%20The%20MA%20Model%20-
%20Completed.ipynb?dl=0
09 Past Values and Past Errors The ARMA Model
10 Modeling Non-Stationary Data The ARIMA
Model
\
prices
data
variables
11 Measuring Volatility The ARCH Model
dataFrame
Of Freedom
Than not
9
075 An ARMA Equivalent of the ARCH Model
12 An ARMA Equivalent of the ARCH The GARCH Model
Next
Variance
variances
11
Component
Have to match
Easier
Yesterday
Choice
31
13 Auto ARIMA
Sections
Include
D
AIC
14 Forecasting
Date ok
Results
Past
The testing set
15 Business Case
