A Blog by
Michael Siebel


I enjoy developing and making things using the tools of statistics and programming. I believe research is only as valuable as how well you are able to communicate it.

U.S. Baby Name Popularity

Michael Siebel
March 2020


Table of Contents

Set Up

In [1]:
# Import Packages
import pandas as pd
import matplotlib.pyplot as pp
import zipfile

# Running a matplotlib style
from jupyterthemes import jtplot
jtplot.style(theme='chesterish')
jtplot.style(context='talk', spines=True, ticks=True, gridlines='')
In [2]:
# Data Load
## read in the names.zip and extract everything
zipfile.ZipFile('names.zip').extractall('.')

Millennial Names, 2014

Let's start by looking at the top 2014 names as an example of common names millennials are giving their babies. Here, we will use raw Pandas syntax (later we will use custom functions).

Top Female Names

In [3]:
names2014 = pd.read_csv('names/yob2014.txt', names=['name','sex','frequency'])
names2014.set_index(['sex']).loc['F'][:10]
Out[3]:
name frequency
sex
F Emma 20799
F Olivia 19674
F Sophia 18490
F Isabella 16950
F Ava 15586
F Mia 13442
F Emily 12562
F Abigail 11985
F Madison 10247
F Charlotte 10048

Top Male Names

In [4]:
names2014.set_index(['sex']).loc['M'][:10]
Out[4]:
name frequency
sex
M Noah 19144
M Liam 18342
M Mason 17092
M Jacob 16712
M William 16687
M Ethan 15619
M Michael 15323
M Alexander 15293
M James 14301
M Daniel 13829

Millennials are naming their baby's Emma, Olivia, Noah, and Liam. My name, Michael, remains number seven on the male list.

Millennial Names, 2008-2014

Tables of Most Popular Female and Male Millennial Names

Custom Functions

Let's create functions to show popular names as tables. This method will enable us to see names from 2008 to 2014 to ensure these millennial baby names were not one-off (i.e., names popular for one year, perhaps due to something occurring in pop culture).

In [5]:
# function to add all years worth of data as a panel data frame
names_all = []
for year in range(1880,2014+1):
    names_all.append(pd.read_csv('names/yob{}.txt'.format(year),
                                 names=['name','sex','frequency']))
    names_all[-1]['year'] = year #append year as the last column

# Save panel data    
allyears = pd.concat(names_all)
allyears_indexed = allyears.set_index(['sex','name','year']).sort_index()

# create a function to do the top ten names for any year based on gender
def topnames(sex, year, limit=10):
    simple = allyears_indexed.loc[sex,:,year].sort_values(by='frequency', ascending=False)
    simple = simple.reset_index()
    simple = simple.drop(['sex','year','frequency'],axis=1).head(limit)
    simple.columns = [year]
    simple.index = simple.index + 1
    return simple

# create a function that gives the top names over a range of years
def top_names_range(sex, start_year, end_year):
    years = [topnames(sex,year) for year in range(start_year,end_year+1)]
    return years[0].join(years[1:])

Top Female Names

In [6]:
top_names_range('F',2008,2014)
Out[6]:
2008 2009 2010 2011 2012 2013 2014
1 Emma Isabella Isabella Sophia Sophia Sophia Emma
2 Isabella Emma Sophia Isabella Emma Emma Olivia
3 Emily Olivia Emma Emma Isabella Olivia Sophia
4 Olivia Sophia Olivia Olivia Olivia Isabella Isabella
5 Ava Ava Ava Ava Ava Ava Ava
6 Madison Emily Emily Emily Emily Mia Mia
7 Sophia Madison Abigail Abigail Abigail Emily Emily
8 Abigail Abigail Madison Madison Mia Abigail Abigail
9 Elizabeth Chloe Chloe Mia Madison Madison Madison
10 Chloe Mia Mia Chloe Elizabeth Elizabeth Charlotte

Top Male Names

In [7]:
top_names_range('M',2008,2014)
Out[7]:
2008 2009 2010 2011 2012 2013 2014
1 Jacob Jacob Jacob Jacob Jacob Noah Noah
2 Michael Ethan Ethan Mason Mason Liam Liam
3 Ethan Michael Michael William Ethan Jacob Mason
4 Joshua Alexander Jayden Jayden Noah Mason Jacob
5 Daniel William William Noah William William William
6 Alexander Joshua Alexander Michael Liam Ethan Ethan
7 William Daniel Noah Ethan Michael Michael Michael
8 Anthony Jayden Daniel Alexander Jayden Alexander Alexander
9 Christopher Noah Aiden Aiden Alexander Jayden James
10 Matthew Christopher Anthony Daniel Aiden Daniel Daniel

Emma and Olivia appear to have been popular for awhile. Noah and Liam appear to be more recent.

Top Male Names the Year I was Born

In [8]:
topnames('M',1988)
Out[8]:
1988
1 Michael
2 Christopher
3 Matthew
4 Joshua
5 Andrew
6 David
7 Justin
8 Daniel
9 James
10 Robert

The year I was born, Michael was the most popular boy's name. In fact, I was one of four Michael's in my 60 student elementary class.

Plotting Millennial Names

In [9]:
# create a function to plot a name across years
def plotname(sex, name):
    data = allyears_indexed.loc[sex,name]
    pp.plot(data.index, data.values)
In [10]:
pp.figure(figsize=(12,3))

females = ['Emma', 'Olivia']
for name in females:
    plotname('F', name)
    
males = ['Noah', 'Liam']
for name in males:
    plotname('M', name)

pp.title('Top Millennial Baby Names')
pp.legend(['Emma', 'Olivia','Noah', 'Liam'])
pp.show()

We can see that Emma was the only name with popularity in the past. Liam is the latest to gain popularity.

Plotting Names of My Family

  • Me: Michael
  • Spouse: Laura
  • Daughter: Elise
  • Son: Emil

First, I will look at each name independently to see their rate of change over time.

In [11]:
plotname('M', 'Michael')

pp.title('Michael')
pp.show()
In [12]:
plotname('F', 'Laura')

pp.title('Laura')
pp.show()
In [13]:
plotname('F', 'Elise')

pp.title('Elise')
pp.show()
In [14]:
plotname('M', 'Emil')

pp.title('Emil')
pp.show()

Next, I will plot the whole family on the same y-axis to show how each name compares in magnitude against one another.

In [15]:
pp.figure(figsize=(12,3))

plotname('M', 'Michael')
plotname('F', 'Laura')
plotname('F', 'Elise')
plotname('M', 'Emil')

pp.title('Siebel Family')
pp.legend(['Michael','Laura','Elise','Emil'])
pp.show()

Finally, what I care about the most, let's look at my children's names on the same y-axis.

In [16]:
pp.figure(figsize=(12,3))

plotname('F', 'Elise')
plotname('M', 'Emil')

pp.title('Siebel Children')
pp.legend(['Elise','Emil'])
pp.show()

What do I draw from this?

This tells me that my daughter may have another Elise in her class - certainly, less likely then how I had three other Michael's in mine.

However, my son is very unlikely to have another Emil in his class.