Class01¶

Introduction to WRDS¶

History¶

Constructed by Wharton in 1993
Written by SAS
To facilitate data collection and data analysis for finance research

Development¶

WRDS cloud and Postgres connection are developed over past few years
More than 350TB data
Over 500 subscribers

WRDS access¶

Register WRDS account

1. Go to WRDS
2. Click REGISTER on top right
3. Fill registration form and then submit
Access methods
- Web interface
- WRDS cloud (SSH)
- Computer language/software (Python, R, SAS, Stata)

Introduction to Python¶

Install Anaconda¶

1. Go to Anaconda

2. Scroll down to find the installation files

Choose the operating system (Linux/MacOS/Windows) and make sure to download the 64-bit Graphical Installer

3. Double click the installation file and follow the wizard to complete the installation

Python editors¶

Jupyter Notebook¶

1. Open your terminal (MacOS) or start Anaconda prompt (Windows), then type in jupyter lab. This will open notebook in your default web browser.

2. Click Python3

3. Type in print('This is Python') in the first cell and press shift + enter (this is the keyboard shortcut to run the code in that cell). The string will be printed out below. In the second cell, let’s write mutiple lines code and run it. In the example below, 2 will be printed.

VSCode¶

There are many other ways to execute Python code. VSCode is a very popular one.

1. Visit VSCode to download installation file and install it.

2. Start VSCode and click the icon on the left bar as indicated below

3. You will find Python extension and click install.

4. Please type in python in the search bar if you cannot see the extension.

5. cmd+n (MacOS) or alt+n (Windows) to create a new file, type in print('This is Python') and save the file with the extension of .py.

6. Check if you have the Anaconda version of Python at left bottom corner. Just click it if you are not sure the Python version. Choose the Python version with anaconda path.

7. cmd+shift+p (MacOS) or ctrl+shift+p to open command palette. Type in setting and click Preferences: Open Setting (UI).

8. Type in send selection in the search on top and make sure Jupyter: Send Selection to Interactive Window is ticked.

9. Move your cursor anythere in the fisrt line and press shift+enter, then an interactive window will be pop up on the right and the result will be printed there. For multiple line code, just select all the codes first and then press shift+enter to run the selection. You will see the results on the right.

Other editors¶

Spyder. This is included in Anaconda distribution (video tutorial)
Vim (an article how to build a Python IDE in Vim)
Emacs (an article how to configure a Python environment in Emacs; Doom Emacs is an out-of-box configuration)

Note

There is a learning curve to use Emacs and Vim if you have no previous experience. Therefore, have a try only if you are interested in Emacs and Vim. Both of them have unique features to make your empirical research more productive. You can also see my Emacs configuration (.emacs.d).

I use Emacs to do all my empirical research.

Retrieve WRDS data¶

WRDS provides Python package wrds to make our life much easier to download data from WRDS (the package is kind of wrapper of postgres connction to WRDS).

Click the binder icon below to lanuch the notebook of class01

Import required packages

import wrds
import pandas as pd

Connect to WRDS

conn = wrds.Connection()

After you run it, type in your WRDS username and press enter; type in your WRDS password and press enter

List databases

lib = conn.list_libraries()
# view first 3 databases
lib[:3]

List tables

# get tables from CRSP database
table = conn.list_tables('crsp')
# view first 3 tables
table[:3]

Describe tables

# list variables in monthly stock file (msf) from CRSP
conn.describe_table('crsp', 'msf')

Extract data

msf = conn.raw_sql("""
    select permno, date, ret
    from crsp.msf
    where date>='01/01/2020' and date<='05/31/2020'
""")

# list first 5 observations
msf.head()

select: claim which variables you want to download and seprate them by comma (,)
from: declare which table the variables are from
where: conditional statement (like a filter)