pandasmore 0.0.1

Extends pandas with common functions used in finance and economics research

pandasmore

The full documentation site is here.

Here is a short description of some of the main functions (see below for more details):

• panel_setup: cleans up dates and panel id’s and sets them as the index (panel id, period date)
• lag: robust lagging that accounts for panel structure, unsorted or duplicate dates, or gaps in the time-series

Install

pip install pandasmore

How to use

First, we set up an example dataset to showcase the functions in this module.

import pandas as pd
import numpy as np
from pandasmore.core import *

df = pd.DataFrame(np.random.rand(8,2), 
                  columns=list('AB'), 
                  index=pd.MultiIndex.from_product(
                      [[1,2],
                       pd.to_datetime(['2010-01','2010-02','2010-02','2010-04']
                                      ).to_period('M')],
                      names=['permno','Mdate']))
df

		A	B
permno	Mdate
1	2010-01	0.065783	0.204804
	2010-02	0.394326	0.129398
	2010-02	0.311802	0.882613
	2010-04	0.573854	0.107920
2	2010-01	0.407374	0.270169
	2010-02	0.345566	0.931718
	2010-02	0.042698	0.989230
	2010-04	0.801888	0.574477

lag(df['A'])

permno  Mdate  
1       2010-01         NaN
        2010-02    0.065783
        2010-02         NaN
        2010-04         NaN
2       2010-01         NaN
        2010-02    0.407374
        2010-02         NaN
        2010-04         NaN
Name: A_lag1, dtype: float64

lag(df['A'], fast=False)

permno  Mdate  
1       2010-01         NaN
        2010-02    0.065783
        2010-02    0.065783
        2010-04         NaN
2       2010-01         NaN
        2010-02    0.407374
        2010-02    0.407374
        2010-04         NaN
Name: A_lag1, dtype: float64