GP02: Explore U.S. Births

The raw data behind the story Some People Are Too Superstitious To Have A Baby On Friday The 13th, which you can read here.

We'll be working with the data set from the Centers for Disease Control and Prevention's National National Center for Health Statistics.

The data set has the following structure:

year - Year
month - Month
date_of_month - Day number of the month
day_of_week - Day of week, where 1 is Monday and 7 is Sunday
births - Number of births

1: Introduction To The Dataset

Lets explore the data and see how it looks.



In [1]:

    
csv_list = open("../data/GP02/US_births_1994-2003_CDC_NCHS.csv").read().split("\n")



In [2]:

    
csv_list[0:10]









    Out[2]:





['year,month,date_of_month,day_of_week,births',
 '1994,1,1,6,8096',
 '1994,1,2,7,7772',
 '1994,1,3,1,10142',
 '1994,1,4,2,11248',
 '1994,1,5,3,11053',
 '1994,1,6,4,11406',
 '1994,1,7,5,11251',
 '1994,1,8,6,8653',
 '1994,1,9,7,7910']

2: Converting Data Into A List Of Lists

The lists needs to be converted to a more structured format to be able to analyze it.



In [3]:

    
def read_csv(filename):
    string_data = open(filename).read()
    string_list = string_data.split("\n")[1:]
    final_list = []
    
    for row in string_list:
        string_fields = row.split(",")
        int_fields = []
        for value in string_fields:
            int_fields.append(int(value))
        final_list.append(int_fields)
    return final_list
        
cdc_list = read_csv("../data/GP02/US_births_1994-2003_CDC_NCHS.csv")



In [4]:

    
cdc_list[0:10]









    Out[4]:





[[1994, 1, 1, 6, 8096],
 [1994, 1, 2, 7, 7772],
 [1994, 1, 3, 1, 10142],
 [1994, 1, 4, 2, 11248],
 [1994, 1, 5, 3, 11053],
 [1994, 1, 6, 4, 11406],
 [1994, 1, 7, 5, 11251],
 [1994, 1, 8, 6, 8653],
 [1994, 1, 9, 7, 7910],
 [1994, 1, 10, 1, 10498]]

3: Calculating Number Of Births Each Month

Now that the data is in a more usable format, we can start to analyze it.



In [5]:

    
def month_births(data):
    births_per_month = {}
    
    for row in data:
        month = row[1]
        births = row[4]
        if month in births_per_month:
            births_per_month[month] = births_per_month[month] + births
        else:
            births_per_month[month] = births
    return births_per_month
    
cdc_month_births = month_births(cdc_list)



In [6]:

    
cdc_month_births









    Out[6]:





{1: 3232517,
 2: 3018140,
 3: 3322069,
 4: 3185314,
 5: 3350907,
 6: 3296530,
 7: 3498783,
 8: 3525858,
 9: 3439698,
 10: 3378814,
 11: 3171647,
 12: 3301860}

4: Calculating Number Of Births Each Day Of Week

Let's now create a function that calculates the total number of births for each unique day of the week.



In [7]:

    
def dow_births(data):
    births_per_dow = {}
    
    for row in data:
        dow = row[3]
        births = row[4]
        if dow in births_per_dow:
            births_per_dow[dow] = births_per_dow[dow] + births
        else:
            births_per_dow[dow] = births
    return births_per_dow
    
cdc_dow_births = dow_births(cdc_list)



In [8]:

    
cdc_dow_births









    Out[8]:





{1: 5789166,
 2: 6446196,
 3: 6322855,
 4: 6288429,
 5: 6233657,
 6: 4562111,
 7: 4079723}

5: Creating A More General Function

It's better to create a single function that works for any column and specify the column we want as a parameter each time we call the function.



In [9]:

    
def calc_counts(data, column):
    sums_dict = {}
    
    for row in data:
        col_value = row[column]
        births = row[4]
        if col_value in sums_dict:
            sums_dict[col_value] = sums_dict[col_value] + births
        else:
            sums_dict[col_value] = births
    return sums_dict

cdc_year_births = calc_counts(cdc_list, 0)
cdc_month_births = calc_counts(cdc_list, 1)
cdc_dom_births = calc_counts(cdc_list, 2)
cdc_dow_births = calc_counts(cdc_list, 3)



In [10]:

    
cdc_year_births









    Out[10]:





{1994: 3952767,
 1995: 3899589,
 1996: 3891494,
 1997: 3880894,
 1998: 3941553,
 1999: 3959417,
 2000: 4058814,
 2001: 4025933,
 2002: 4021726,
 2003: 4089950}



In [11]:

    
cdc_month_births









    Out[11]:





{1: 3232517,
 2: 3018140,
 3: 3322069,
 4: 3185314,
 5: 3350907,
 6: 3296530,
 7: 3498783,
 8: 3525858,
 9: 3439698,
 10: 3378814,
 11: 3171647,
 12: 3301860}



In [12]:

    
cdc_dom_births









    Out[12]:





{1: 1276557,
 2: 1288739,
 3: 1304499,
 4: 1288154,
 5: 1299953,
 6: 1304474,
 7: 1310459,
 8: 1312297,
 9: 1303292,
 10: 1320764,
 11: 1314361,
 12: 1318437,
 13: 1277684,
 14: 1320153,
 15: 1319171,
 16: 1315192,
 17: 1324953,
 18: 1326855,
 19: 1318727,
 20: 1324821,
 21: 1322897,
 22: 1317381,
 23: 1293290,
 24: 1288083,
 25: 1272116,
 26: 1284796,
 27: 1294395,
 28: 1307685,
 29: 1223161,
 30: 1202095,
 31: 746696}



In [13]:

    
cdc_dow_births









    Out[13]:





{1: 5789166,
 2: 6446196,
 3: 6322855,
 4: 6288429,
 5: 6233657,
 6: 4562111,
 7: 4079723}