pandasmore 0.0.2

Extends pandas with common functions used in finance and economics research

pandasmore

The full documentation site is here, and the GitHub page is here.

Here is a short description of some of the main functions (more details below and in the documentation):

• setup_tseries: cleans up dates and sets them as the index
• setup_panel: 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
import pandasmore as pdr

raw = pd.DataFrame(np.random.rand(15,2), 
                    columns=list('AB'), 
                    index=pd.MultiIndex.from_product(
                        [[1,2, np.nan],[np.nan,'2010-01','2010-02','2010-02','2010-04']],
                        names = ['firm_id','date'])
                      ).reset_index()
raw

	firm_id	date	A	B
0	1.0	NaN	0.943132	0.981995
1	1.0	2010-01	0.328816	0.473158
2	1.0	2010-02	0.177921	0.835497
3	1.0	2010-02	0.928199	0.743025
4	1.0	2010-04	0.857208	0.742693
5	2.0	NaN	0.147470	0.357477
6	2.0	2010-01	0.172676	0.978518
7	2.0	2010-02	0.391758	0.574734
8	2.0	2010-02	0.824737	0.863340
9	2.0	2010-04	0.847638	0.293925
10	NaN	NaN	0.969513	0.842419
11	NaN	2010-01	0.491236	0.194837
12	NaN	2010-02	0.854151	0.267796
13	NaN	2010-02	0.461259	0.010185
14	NaN	2010-04	0.735704	0.601400

df = pdr.setup_tseries(raw.query('firm_id==1'),
                        time_var='date', time_var_format="%Y-%m",
                        freq='M')
df

	date	dtdate	firm_id	A	B
Mdate
2010-01	2010-01	2010-01-01	1.0	0.328816	0.473158
2010-02	2010-02	2010-02-01	1.0	0.928199	0.743025
2010-04	2010-04	2010-04-01	1.0	0.857208	0.742693

df = pdr.setup_panel(raw,
                        panel_ids='firm_id',
                        time_var='date', time_var_format="%Y-%m",
                        freq='M')
df

		date	dtdate	A	B
firm_id	Mdate
1	2010-01	2010-01	2010-01-01	0.328816	0.473158
	2010-02	2010-02	2010-02-01	0.928199	0.743025
	2010-04	2010-04	2010-04-01	0.857208	0.742693
2	2010-01	2010-01	2010-01-01	0.172676	0.978518
	2010-02	2010-02	2010-02-01	0.824737	0.863340
	2010-04	2010-04	2010-04-01	0.847638	0.293925

pdr.lag(df['A'])

permno  Mdate  
1       2010-01         NaN
        2010-02    0.698770
        2010-04         NaN
2       2010-01         NaN
        2010-02    0.834091
        2010-04         NaN
Name: A_lag1, dtype: float64