capstone_buoys.py

# coding: utf-8

# # Analyzing data from the National Bouy System

# ## Background
# 
# The National Data Buoys Data center publish oceanographic data from their buoys on their web. Bouys send up their measurements every hour and the data is published as a spreadsheet in the NOAA web site.
# 
# We want to collect the data for any, or many, bouys, clean it up, and analyze it to make a plot.
# 
# ## Challenges
# 
# - Get the data from the web and read it into our Python notebook.
# - Clean up the data, which sometimes has missing measurements.
# - Choose the bouys we want to analyze using their ID numbers.
# - Run our analysis (we will just make a plot).
# - Save our code as a script or module for reusing it later.

# ## References 
# 
# - National Data Buoy Center: http://www.ndbc.noaa.gov/
# - Data for Station 44255 - NE Burgeo Bank: http://www.ndbc.noaa.gov/station_realtime.php?station=44255
# - How a buoy get an ID: http://www.ndbc.noaa.gov/staid.shtml
# - All the station IDs: http://www.ndbc.noaa.gov/to_station.shtml

# In[96]:

#So convert into 
import StringIO #For converting into strings   #Moving earlier to allow for importing all buoys.txt

import pandas as pd #cf. Excel. Go-to for data processing #as pd is just giving shortcut

import requests #Library for getting stuff from web

import pylab  

import numpy as np
from numpy import nan

import matplotlib.pyplot as plt


# In[82]:

# Retrieve data
# url = 'http://www.ndbc.noaa.gov/data/realtime2/44255.txt'
url = 'http://www.ndbc.noaa.gov/view_text_file.php?filename=44011h2012.txt.gz&dir=data/historical/stdmet/'
response = requests.get(url)


# In[83]:

# type(response.text)


# In[84]:

data_str = StringIO.StringIO(response.text)


# In[85]:

data = pd.read_csv(StringIO.StringIO(response.text), 
                   delim_whitespace=True, 
                   skiprows=[1,2], 
                   usecols=[0,1,2,3,6,8])


# In[126]:

# import numpy as np
# from numpy import nan #moved to top
data=data.replace('MM', nan)
data=data.replace('99', nan)
data=data.dropna(axis=0)

z = np.polyfit(data.WSPD, data.WVHT, 2)
p = np.poly1d(z)
wspd=range(20)
pylab.plot(wspd,p(wspd))

# the line equation:
# print “y=%.6fx+(%.6f)”%(z[0],z[1])


# import scipy.signal
# detrended=scipy.signal.detrend(data.WSPD, data.WVHT)


# In[110]:

get_ipython().magic(u'matplotlib inline')

plt.plot(data.WSPD, data.WVHT, '*')

# data.WSPD.plot()


# In[88]:

print data.WSPD.mean()


# In[89]:

#nbviewer.com


# In[90]:

import capstone_buoys #after saving to cd as .py