Analyzing Baby names with Pandas

import matplotlib.pyplot as plt
import matplotlib.style as sty
import os
import re
import pandas as pd
import numpy as np
import matplotlib
from sys import exit
#an easy way to get all the imports installed is by just installing Anaconda (especially true for windows users)


class BabyNameAnalyzer:
    """Anaylze SSA baby name HTML files -- Use with a web scraper for maximum effectiveness get it here http://pastebin.com/9HsU767u"""

    def __init__(self):
        sty.use('ggplot')
        self.namedict = {}
        self.boyrankings = {}
        self.girlrankings = {}
        self.extractNames()
        self.organizeNames()

    def extractNames(self):
        """Extract the names from the Python Google class html files

        (in this example run the program from that directory)
        """
        for root, dirs, files in os.walk(os.path.dirname(os.path.realpath(__file__))):
            patt = re.compile(r'<tr align="right">\s*<td>(\d{1,4})<\/td>\s*<td>([A-Z][a-z]+?)<\/td>\s*<td>([A-Z][a-z]+?)<\/td>')
            for file in files:
                year = file[4:-5]
                self.namedict[year] = []
                if file[-4:] == "html":
                    myfile = open(os.path.join(root, file))
                    htmlText = myfile.read()
                    found = re.findall(patt,htmlText)
                    myfile.close()
                    [self.namedict[year].append(find) for find in found]

    def organizeNames(self):
        """Organize the names into Python dictionary Objects for both girls and boys"""
        for key in self.namedict.keys():
            for nametuple in self.namedict[key]:
                if nametuple[1] in self.boyrankings and nametuple[2] in self.girlrankings:
                    self.boyrankings[nametuple[1]].append((int(key), int(nametuple[0])))
                    self.girlrankings[nametuple[2]].append((int(key), int(nametuple[0])))
                elif nametuple[1] in self.boyrankings and nametuple[2] not in self.girlrankings:
                    self.boyrankings[nametuple[1]].append((int(key), int(nametuple[0])))
                    self.girlrankings[nametuple[2]] = [(int(key), int(nametuple[0])),]
                elif nametuple[1] not in self.boyrankings and nametuple[2] in self.girlrankings:
                    self.boyrankings[nametuple[1]] = [(int(key), int(nametuple[0])),]
                    self.girlrankings[nametuple[2]].append((int(key), int(nametuple[0])))
                else:
                    self.boyrankings[nametuple[1]] = [(int(key), int(nametuple[0])),]
                    self.girlrankings[nametuple[2]] = [(int(key), int(nametuple[0])),]


    def plotName(self, gender, name):
        """Given a gender and name create a matplotlib line graph mapping the rankings over the years"""
        if gender == "boy":
            data = self.boyrankings[name]
            data.sort(key=lambda x: x[0])
            df = pd.DataFrame(data, columns=['Year', 'Rank'])
            dfplot = df.plot(figsize=(20,6), grid=True, kind='line', x='Year', y='Rank', title="Rankings for {}".format(name), xticks=[x for x in range(1880,2016) if x % 4 ==0])
            plt.gca().invert_yaxis()
            figure = dfplot.get_figure()
            figure.savefig("{}.png".format(name))
        elif gender == "girl":
            data = self.girlrankings[name]
            data.sort(key=lambda x: x[0])
            df = pd.DataFrame(data, columns=['Year', 'Rank'])
            dfplot = df.plot(figsize=(20,6), grid=True, kind='line', x='Year', y='Rank', title="Rankings for {}".format(name), xticks=[x for x in range(1880,2016) if x % 4 ==0])
            plt.gca().invert_yaxis()
            figure = dfplot.get_figure()
            figure.savefig("{}.png".format(name))
        else:
            pass

if __name__ == '__main__':
  main()