SF.net SVN: matplotlib: [3972] trunk/py4science/examples

[<jd...@us...>]

Revision: 3972
          http://matplotlib.svn.sourceforge.net/matplotlib/?rev=3972&view=rev
Author:   jdh2358
Date:     2007-10-20 14:23:56 -0700 (Sat, 20 Oct 2007)

Log Message:
-----------
added stock records CSV example

Added Paths:
-----------
    trunk/py4science/examples/skel/stock_records_skel.py
    trunk/py4science/examples/stock_records.py

Added: trunk/py4science/examples/skel/stock_records_skel.py
===================================================================
--- trunk/py4science/examples/skel/stock_records_skel.py	                        (rev 0)
+++ trunk/py4science/examples/skel/stock_records_skel.py	2007-10-20 21:23:56 UTC (rev 3972)
@@ -0,0 +1,71 @@
+"""
+Download historical pricing record arrays for a universe of stocks
+from Yahoo Finance using urllib.  Load them into numpy record arrays
+using matplotlib.mlab.csv2rec, and do some batch processing -- make
+date vs price charts for each one, and compute the return since 2003
+for each stock.  Sort the returns and print out the tickers of the 4
+biggest winners
+"""
+import os, datetime, urllib
+import matplotlib.mlab as mlab  # contains csv2rec
+import numpy as npy
+import pylab as p
+
+def fetch_stock(ticker):
+    """
+    download the CSV file for stock with ticker and return a numpy
+    record array.  Save the CSV file as TICKER.csv where TICKER is the
+    stock's ticker symbol.
+
+    Extra credit for supporting a start date and end date, and
+    checking to see if the file already exists on the local file
+    system before re-downloading it
+    """
+    fname = '%s.csv'%ticker
+    url = XXX # create the url for this ticker
+
+    # the os.path module contains function for checking whether a file
+    # exists, and fetch it if not
+    XXX
+
+    # load the CSV file intoo a numpy record array
+    r = XXX
+
+    # note that the CSV file is sorted most recent date first, so you
+    # will probably want to sort the record array so most recent date
+    # is last
+    XXX
+    return r
+
+tickers = 'INTC', 'MSFT', 'YHOO', 'GOOG', 'GE', 'WMT', 'AAPL'
+
+# we want to compute returns since 2003, so define the start date as a
+# datetime.datetime instance
+startdate = XXX
+
+# we'll store a list of each return and ticker for analysis later
+data = []   # a list of (return, ticker) for each stock 
+fig = p.figure()
+for ticker in tickers:
+    print 'fetching', ticker
+    r = fetch_stock(ticker)
+    
+    # select the numpy records where r.date>=startdatre use numpy mask
+    # indexing to restrict r to just the dates > startdate
+    r = XXX
+    price = XXX   # set price equal to the adjusted close
+    returns = XXX # return is the (price-p0)/p0
+    XXX           # store the data
+
+    # plot the returns by date for each stock using pylab.plot, adding
+    # a label for the legend
+    XXX
+
+# use pylab legend command to build a legend
+XXX
+
+# now sort the data by returns and print the results for each stock
+XXX
+
+# show the figures
+p.show()

Added: trunk/py4science/examples/stock_records.py
===================================================================
--- trunk/py4science/examples/stock_records.py	                        (rev 0)
+++ trunk/py4science/examples/stock_records.py	2007-10-20 21:23:56 UTC (rev 3972)
@@ -0,0 +1,72 @@
+"""
+Download historical pricing record arrays for a universe of stocks
+from Yahoo Finance using urllib.  Load them into numpy record arrays
+using matplotlib.mlab.csv2rec, and do some batch processing -- make
+date vs price charts for each one, and compute the return since 2003
+for each stock.  Sort the returns and print out the tickers of the 4
+biggest winners
+"""
+import os, datetime, urllib
+import matplotlib.mlab as mlab  # contains csv2rec
+import numpy as npy
+import pylab as p
+
+def fetch_stock(ticker):
+    """
+    download the CSV file for stock with ticker and return a numpy
+    record array.  Save the CSV file as TICKER.csv where TICKER is the
+    stock's ticker symbol.
+
+    Extra credit for supporting a start date and end date, and
+    checking to see if the file already exists on the local file
+    system before re-downloading it
+    """
+    fname = '%s.csv'%ticker
+    url = 'http://ichart.finance.yahoo.com/table.csv?' +\
+   's=%s&d=9&e=20&f=2007&g=d&a=0&b=29&c=1993&ignore=.csv'%ticker
+
+    # the os.path module contains function for checking whether a file
+    # exists
+    if not os.path.exists(fname):
+        urllib.urlretrieve(url, fname)
+    r = mlab.csv2rec(fname)
+
+    # note that the CSV file is sorted most recent date first, so you
+    # will probably want to sort the record array so most recent date
+    # is last
+    r.sort()
+    return r
+
+tickers = 'INTC', 'MSFT', 'YHOO', 'GOOG', 'GE', 'WMT', 'AAPL'
+
+# we want to compute returns since 2003, so define the start date
+startdate = datetime.datetime(2003,1,1)
+
+# we'll store a list of each return and ticker for analysis later
+data = []   # a list of (return, ticker) for each stock 
+fig = p.figure()
+for ticker in tickers:
+    print 'fetching', ticker
+    r = fetch_stock(ticker)
+    
+    # select the numpy records where r.date>=startdatre
+
+    r = r[r.date>=startdate]
+    price = r.adj_close                 # set price equal to the adjusted close
+    returns = (price-price[0])/price[0] # return is the (price-p0)/p0
+    data.append((returns[-1], ticker))  # store the data
+
+    # plot the returns by date for each stock
+    p.plot(r.date, returns, label=ticker)
+
+p.legend(loc='upper left')
+
+# now sort the data by returns and print the results for each stock
+data.sort()
+for g, ticker in data:
+    print '%s: %1.1f%%'%(ticker, 100*g)
+
+
+p.savefig('fig/stock_records.png', dpi=100)
+p.savefig('fig/stock_records.eps')
+p.show()


This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.