Download history stock prices automatically from yahoo finance in python

Question

Is there a way to automatically download historical prices of stocks from yahoo finance or google finance (csv format)? Preferably in Python.

Answer 1

When you're going to work with such time series in Python, pandas is indispensable. And here's the good news: it comes with a historical data downloader for Yahoo: pandas.io.data.DataReader.

from pandas.io.data import DataReader
from datetime import datetime

ibm = DataReader('IBM',  'yahoo', datetime(2000, 1, 1), datetime(2012, 1, 1))
print(ibm['Adj Close'])

Here's an example from the pandas documentation.

Update for pandas >= 0.19:

The pandas.io.data module has been removed from pandas>=0.19 onwards. Instead, you should use the separate pandas-datareader package. Install with:

pip install pandas-datareader

And then you can do this in Python:

import pandas_datareader as pdr
from datetime import datetime

ibm = pdr.get_data_yahoo(symbols='IBM', start=datetime(2000, 1, 1), end=datetime(2012, 1, 1))
print(ibm['Adj Close'])

Downloading from Google Finance is also supported.

There's more in the documentation of pandas-datareader.

Answer 2

Short answer: Yes. Use Python's urllib to pull the historical data pages for the stocks you want. Go with Yahoo! Finance; Google is both less reliable, has less data coverage, and is more restrictive in how you can use it once you have it. Also, I believe Google specifically prohibits you from scraping the data in their ToS.

Longer answer: This is the script I use to pull all the historical data on a particular company. It pulls the historical data page for a particular ticker symbol, then saves it to a csv file named by that symbol. You'll have to provide your own list of ticker symbols that you want to pull.

import urllib

base_url = "http://ichart.finance.yahoo.com/table.csv?s="
def make_url(ticker_symbol):
    return base_url + ticker_symbol

output_path = "C:/path/to/output/directory"
def make_filename(ticker_symbol, directory="S&P"):
    return output_path + "/" + directory + "/" + ticker_symbol + ".csv"

def pull_historical_data(ticker_symbol, directory="S&P"):
    try:
        urllib.urlretrieve(make_url(ticker_symbol), make_filename(ticker_symbol, directory))
    except urllib.ContentTooShortError as e:
        outfile = open(make_filename(ticker_symbol, directory), "w")
        outfile.write(e.content)
        outfile.close()

Answer 3

Extending @Def_Os's answer with an actual demo...

As @Def_Os has already said - using Pandas Datareader makes this task a real fun

In [12]: from pandas_datareader import data

pulling all available historical data for AAPL starting from 1980-01-01

#In [13]: aapl = data.DataReader('AAPL', 'yahoo', '1980-01-01')

# yahoo api is inconsistent for getting historical data, please use google instead.
In [13]: aapl = data.DataReader('AAPL', 'google', '1980-01-01')

first 5 rows

In [14]: aapl.head()
Out[14]:
                 Open       High     Low   Close     Volume  Adj Close
Date
1980-12-12  28.750000  28.875000  28.750  28.750  117258400   0.431358
1980-12-15  27.375001  27.375001  27.250  27.250   43971200   0.408852
1980-12-16  25.375000  25.375000  25.250  25.250   26432000   0.378845
1980-12-17  25.875000  25.999999  25.875  25.875   21610400   0.388222
1980-12-18  26.625000  26.750000  26.625  26.625   18362400   0.399475

last 5 rows

In [15]: aapl.tail()
Out[15]:
                 Open       High        Low      Close    Volume  Adj Close
Date
2016-06-07  99.250000  99.870003  98.959999  99.029999  22366400  99.029999
2016-06-08  99.019997  99.559998  98.680000  98.940002  20812700  98.940002
2016-06-09  98.500000  99.989998  98.459999  99.650002  26419600  99.650002
2016-06-10  98.529999  99.349998  98.480003  98.830002  31462100  98.830002
2016-06-13  98.690002  99.120003  97.099998  97.339996  37612900  97.339996

save all data as CSV file

In [16]: aapl.to_csv('d:/temp/aapl_data.csv')

d:/temp/aapl_data.csv - 5 first rows

Date,Open,High,Low,Close,Volume,Adj Close
1980-12-12,28.75,28.875,28.75,28.75,117258400,0.431358
1980-12-15,27.375001,27.375001,27.25,27.25,43971200,0.408852
1980-12-16,25.375,25.375,25.25,25.25,26432000,0.378845
1980-12-17,25.875,25.999999,25.875,25.875,21610400,0.38822199999999996
1980-12-18,26.625,26.75,26.625,26.625,18362400,0.399475
...

Answer 4

There is already a library in Python called yahoo_finance so you'll need to download the library first using the following command line:

sudo pip install yahoo_finance

Then once you've installed the yahoo_finance library, here's a sample code that will download the data you need from Yahoo Finance:

#!/usr/bin/python
import yahoo_finance
import pandas as pd

symbol = yahoo_finance.Share("GOOG")
google_data = symbol.get_historical("1999-01-01", "2016-06-30")
google_df = pd.DataFrame(google_data)

# Output data into CSV
google_df.to_csv("/home/username/google_stock_data.csv")

This should do it. Let me know if it works.

UPDATE: The yahoo_finance library is no longer supported.

Answer 5

You can check out the yahoo_fin package. It was initially created after Yahoo Finance changed their API (documentation is here: http://theautomatic.net/yahoo_fin-documentation).

from yahoo_fin import stock_info as si

aapl_data = si.get_data("aapl")

nflx_data = si.get_data("nflx")

aapl_data.head()

nflx_data.head()

aapl_data.to_csv("aapl_data.csv")

nflx_data.to_csv("nflx_data.csv")

Answer 6

It's trivial when you know how:

import yfinance as yf
df = yf.download('CVS', '2015-01-01')
df.to_csv('cvs-health-corp.csv')

If you wish to plot it:

import finplot as fplt
fplt.candlestick_ochl(df[['Open','Close','High','Low']])
fplt.show()