survey.py

"""This file contains code for use with "Think Stats",
by Allen B. Downey, available from greenteapress.com

Copyright 2010 Allen B. Downey
License: GNU GPLv3 http://www.gnu.org/licenses/gpl.html
"""

import sys
import gzip
import os

class Record(object):
    """Represents a record."""

class Respondent(Record): 
    """Represents a respondent."""

class Pregnancy(Record):
    """Represents a pregnancy."""

class Table(object):
    """Represents a table as a list of objects"""

    def __init__(self):
        self.records = []
        
    def __len__(self):
        return len(self.records)

    def ReadFile(self, data_dir, filename, fields, constructor, n=None):
        """Reads a compressed data file builds one object per record.

        Args:
            data_dir: string directory name
            filename: string name of the file to read

            fields: sequence of (name, start, end, case) tuples specifying 
            the fields to extract

            constructor: what kind of object to create
        """
        filename = os.path.join(data_dir, filename)

        if filename.endswith('gz'):
            fp = gzip.open(filename)
        else:
            fp = open(filename)

        for i, line in enumerate(fp):
            if i == n:
                break
            record = self.MakeRecord(line, fields, constructor)
            self.AddRecord(record)
        fp.close()

    def MakeRecord(self, line, fields, constructor):
        """Scans a line and returns an object with the appropriate fields.

        Args:
            line: string line from a data file

            fields: sequence of (name, start, end, cast) tuples specifying 
            the fields to extract

            constructor: callable that makes an object for the record.

        Returns:
            Record with appropriate fields.
        """
        obj = constructor()
        for (field, start, end, cast) in fields:
            try:
                s = line[start-1:end]
                val = cast(s)
            except ValueError:
                #print line
                #print field, start, end, s
                val = 'NA'
            setattr(obj, field, val)
        return obj

    def AddRecord(self, record):
        """Adds a record to this table.

        Args:
            record: an object of one of the record types.
        """
        self.records.append(record)

    def ExtendRecords(self, records):
        """Adds records to this table.

        Args:
            records: a sequence of record object
        """
        self.records.extend(records)

    def Recode(self):
        """Child classes can override this to recode values."""
        pass


class Respondents(Table):
    """Represents the respondent table."""

    def ReadRecords(self, data_dir='.', n=None):
        filename = self.GetFilename()
        self.ReadFile(data_dir, filename, self.GetFields(), Respondent, n)
        self.Recode()

    def GetFilename(self):
        return '2002FemResp.dat.gz'

    def GetFields(self):
        """Returns a tuple specifying the fields to extract.

        The elements of the tuple are field, start, end, case.

                field is the name of the variable
                start and end are the indices as specified in the NSFG docs
                cast is a callable that converts the result to int, float, etc.
        """
        return [
            ('caseid', 1, 12, int),
            ]

class Pregnancies(Table):
    """Contains survey data about a Pregnancy."""

    def ReadRecords(self, data_dir='.', n=None):
        filename = self.GetFilename()
        self.ReadFile(data_dir, filename, self.GetFields(), Pregnancy, n)
        self.Recode()

    def GetFilename(self):
        return '2002FemPreg.dat.gz'

    def GetFields(self):
        """Gets information about the fields to extract from the survey data.

        Documentation of the fields for Cycle 6 is at
        http://nsfg.icpsr.umich.edu/cocoon/WebDocs/NSFG/public/index.htm

        Returns:
            sequence of (name, start, end, type) tuples
        """
        return [
            ('caseid', 1, 12, int),
            ('nbrnaliv', 22, 22, int),
            ('babysex', 56, 56, int),
            ('birthwgt_lb', 57, 58, int),
            ('birthwgt_oz', 59, 60, int),
            ('prglength', 275, 276, int),
            ('outcome', 277, 277, int),
            ('birthord', 278, 279, int),
            ('agepreg', 284, 287, int),
            ('finalwgt', 423, 440, float),
            ]

    def Recode(self):
        for rec in self.records:

            # divide mother's age by 100
            try:
                if rec.agepreg != 'NA':
                    rec.agepreg /= 100.0
            except AttributeError:
                pass

            # convert weight at birth from lbs/oz to total ounces
            # note: there are some very low birthweights
            # that are almost certainly errors, but for now I am not
            # filtering
            try:
                if (rec.birthwgt_lb != 'NA' and rec.birthwgt_lb < 20 and
                    rec.birthwgt_oz != 'NA' and rec.birthwgt_oz <= 16):
                    rec.totalwgt_oz = rec.birthwgt_lb * 16 + rec.birthwgt_oz
                else:
                    rec.totalwgt_oz = 'NA'
            except AttributeError:
                pass


def main(name, data_dir='.'):
    resp = Respondents()
    resp.ReadRecords(data_dir)
    print 'Number of respondents', len(resp.records)

    preg = Pregnancies()
    preg.ReadRecords(data_dir)
    print 'Number of pregnancies', len(preg.records)

    
if __name__ == '__main__':
    main(*sys.argv)