Machine Learning with Python 1st Edition Oliver Theobald download
Machine Learning with Python 1st Edition Oliver Theobald download
https://ebookname.com/product/machine-learning-with-python-1st-
edition-oliver-theobald/
https://ebookname.com/product/machine-learning-with-spark-and-
python-2nd-edition-michael-bowles/
https://ebookname.com/product/python-machine-learning-
cookbook-1st-edition-joshi/
https://ebookname.com/product/python-machine-learning-1st-
edition-wei-meng-lee/
https://ebookname.com/product/cd-and-dvd-forensics-1st-edition-
paul-crowley/
The English Medieval Town A Reader in English Urban
History 1200 1540 Richard Holt
https://ebookname.com/product/the-english-medieval-town-a-reader-
in-english-urban-history-1200-1540-richard-holt/
https://ebookname.com/product/the-cambridge-companion-to-adam-
smith-first-edition-knud-haakonssen/
https://ebookname.com/product/principles-for-best-practice-in-
clinical-audit-1st-edition-national-institute-for-clinical-
excellence/
https://ebookname.com/product/contemporary-colonialities-in-
mexico-and-beyond-1st-edition-kathleen-ann-myers/
https://ebookname.com/product/bloodshot-cherie-priest/
Global Citizenship A Critical Introduction 1st Edition
Nigel Dower (Editor)
https://ebookname.com/product/global-citizenship-a-critical-
introduction-1st-edition-nigel-dower-editor/
First Edition
www.scatterplotpress.com
2
FIND US ON:
Newsletter
http://eepurl.com/gKjQij
Enjoy book recommendations, free giveaways of future book releases from the
author, and other blog posts and news concerning machine learning, trends, and
data science.
Teachable
http://scatterplotpress.com
Skillshare
www.skillshare.com/user/machinelearning_beginners
For introductory video courses on machine learning and video lessons from other
instructors.
Instagram
machinelearning_beginners
3
TABLE OF CONTENTS
FOREWORD ..................................................................................................................... 5
INTRODUCTION ............................................................................................................. 10
4
FOREWORD
While it’s luring to see trends rise quickly, it’s important to see long periods of
resilience before the curve. For those pursuing a career in machine learning, it’s
reassuring to know this field of study not only predates the Internet and the moon
landing but also most readers of this book.
Machine learning is not an overnight movement and the path to the present day
has been anything but smooth sailing. Conceptual theories emerged in the 1950s
but progress was stalled by computational constraints and limited data. This
resulted in a logjam of research and good intentions as theoretical models of
prediction, algorithm design, and extrapolation of future possibilities accumulated
in research institutions until powerful processing chips and large datasets emerged
in the 1990s. Renewed interest helped to breach the gap between theory and
capability during this decade but it still wasn’t enough to push field-altering
breakthroughs in the space of deep learning.
That breakthrough came in 2009 when Adjunct Professor Andrew Ng and his team
at Stanford University experimented with tethering gaming chips—better known for
image rendering—to solve complex data problems. The combination of inexpensive
GPU (graphic processing unit) chips and compute-intensive algorithms pushed the
lead domino in the development of deep learning. This crucial breakthrough
coalesced with other developments in reinforcement learning to spark a surge in
interest, an oversupply of newspaper analogies to Hollywood movies, and an
international hunt for AI talent.
In 2016, media interest climbed to a new high at the glitzy Four Seasons Hotel in
Seoul, where TV cameras locked lenses on an 18-by-18 Go board with the world
champion on one side and an AI program on the other. The game of Go consists of
billions of permutations and commentators described the then world champion, Lee
Sodol, as having a sixth sense for interpreting the state of play. His opponent was
AlphaGo, a sophisticated deep learning model designed to outperform any
opponent—mortal or synthetic.
The team of human developers responsible for designing the AlphaGo program
scarcely knew the rules of the game when they began work on the project, but
they watched on excitedly as AlphaGo performed its first move.
The AI model unsettled Lee early—forcing him to take a nervous cigarette break—
before systemically defeating the South Korean four games to one. News headlines
of AlphaGo’s cold and mechanistic victory beamed across the globe—as had been
the case with other televised AI feats before it. Predictably, these reports focused
on the superiority of machine intelligence over humans.
Contrary to these initial headlines, the 2017 Netflix documentary AlphaGo helped
to later realign attention towards the human ingenuity behind AlphaGo’s victory.
The documentary details the lead-up to Seoul and in doing so shines the light on a
team of talented employees thriving in a new and far-reaching line of work.
Dressed in casual attire, the AlphaGo team can be seen working hard behind their
screens stocking the model with training data, optimizing its hyperparameters, and
5
coordinating vital computational resources before extracting game tactics from
human experts honed over many years of competition.
Despite its prolific success, the AlphaGo program has not replaced any of the
programmers who worked on its source code or taken away their salaries. In fact,
the development of AlphaGo has helped to expand the size and profile of the
company DeepMind Technologies, which was acquired by Alphabet Inc earlier in
2014.
Working in AI
After two AI winters and ongoing battles for academic funding, we have entered a
golden age in industry employment. Complex databases, fast and affordable
processing units, and advanced algorithms have rejuvenated established fields of
human expertise in mathematics, statistics, computer programming, graphics and
visualization as well as good old problem-solving skills.
In a global job market steadily automated and simplified by Web 2.0 technology,
the field of machine learning provides a professional nirvana for human ingenuity
and meaningful work. It’s a cognitively demanding occupation; one that goes far
beyond tuning ad campaigns or tracking web traffic on side-by-side monitors. With
jobs in this industry demanding expertise in three distinct fields, achieving machine
intelligence is far from easy and demands a high level of expertise.
The ideal skillset for a machine learning developer spans coding, data management,
and knowledge of statistics and mathematics. Optional areas of expertise include
data visualization, big data management, and practical experience in distributed
computing architecture. This book converges on the vital coding part of machine
learning using Python.
Released in 1991 by Guido van Rossum, Python is widely used in the field of
machine learning and is easy to learn courtesy of van Rossum’s emphasis on code
readability. Python is versatile too; while other popular languages like R offer
advantages in advanced mathematical operations and statistical functions, they
offer limited practical use outside of hard data crunching. The utility of Python,
however, extends to data collection (web scraping) and data piping (Hadoop and
Spark), which are important for sending data to the operating table. In addition,
Python is convertible to C and C++, enabling practitioners to run code on graphic
processing units reserved for advanced computation.
The other advantages of learning a popular programming language (such as Python)
are the depth of jobs and the spread of relevant support. Access to documentation,
tutorials, and assistance from a helpful community to troubleshoot code problems
cannot be overlooked and especially for anyone beginning their journey in the
complex world of computer programming.
As a practical introduction to coding machine learning models, this book falls short
of a complete introduction to programming with Python. Instead, general nuances
are explained to enlighten beginners without stalling the progress of experienced
programmers. For those new to Python, a basic overview of Python can be found
in the Appendix section of this book. It’s also recommended that you spend 2-3
6
hours watching introductory Python tutorials on YouTube or Udemy if this is your
first time working with Python.
7
DATASETS USED IN THIS BOOK
For any issues accessing and downloading these three datasets, please
contact the author at oliver.theobald@scatterplotpress.com
Advertising Dataset
Overview: This dataset contains fabricated information about the features of users
responding to online advertisements, including their sex, age, location, daily time
spent online, and whether they clicked on the target advertisement. The dataset
was created by Udemy course instructor Jose Portilla of Pierian Data and is used in
his course Python for Data Science and Machine Learning Bootcamp.
Features: 10
Missing values: No
File name: advertising.csv
http://scatterplotpress.com/p/datasets
https://www.kaggle.com/anthonypino/melbourne-housing-market/#Melbourne_housing_FULL.csv
http://scatterplotpress.com/p/datasets
Kickstarter Dataset
Overview: Kickstarter.com is the world's largest crowd-funding platform for
creative projects and this dataset was created using data extracted from the
Kickstarter website.
Features: 35
8
Contains missing values: Yes
File name: 18k_Projects.csv
https://www.kaggle.com/tayoaki/kickstarter-dataset
9
1
INTRODUCTION
As an empirical and specialized field of data science and a dominant sub-field of AI,
machine learning1 describes the ability of computer models to learn from data and
perform cognitive reasoning without direct programming.2 This is a process known
as self-learning—an exciting but somewhat vague concept that underpins machine
learning. While the human programmer maintains ownership of variable selection
and setting algorithm learning hyperparameters (settings), the decision model
interprets patterns and generates an output without a direct command. This course
of action serves as a major distinction from traditional computer programming
where computers are designed to produce fixed outputs in response to pre-
programmed commands.
The initial blueprints for machine learning were conceived by Arthur Samuel while
working for IBM as an engineer in the late 1950s. Samuel defined machine
learning as a subfield of computer science that provides computers the ability to
learn without being explicitly programmed.3 Incorporating probability theory and
statistical modeling, Samuel outlined the potential for machines to detect patterns
and improve performance based on data and empirical information; all without
direct programming commands.
Samuel held that by using data as input, machines could mimic the ability of
humans to learn and identify optimal decisions without explicit code commands
from the programmer. While human programmers were required to facilitate the
input of data and the selection of algorithm(s), they would forego the role of rule-
maker under Samuel’s radical new theory.
In 1959, Samuel published a paper in the IBM Journal of Research and
Development investigating the application of machine learning in the game of
checkers. The goal of his research was to program a computer to gradually exceed
the capabilities of the person who programmed it. The machine was designed to
assess the state of a checkers board and incorporated probability theory to identify
a move that would best lead to a winning outcome. After each game, the program
integrated experience and logged new strategies to refine its performance for play
against the next opponent. This process repeated until the computer program was
able to consistently beat its programmer.
Samuel’s chess program held a competitive edge over symbolic systems that were
in vogue at this time and which relied on pre-programmed knowledge. Unlike
1
The word “machine” was a common byname for computers during this time and the moniker has stuck
over the decades.
2
Aurélien Géron, “Hands-On Machine Learning with Scikit-Learn and TensorFlow: Concepts, Tools, and
Techniques to Build Intelligent Systems,” O’Reilly Media, 2017.
3
Arthur Samuel, “Some Studies in Machine Learning Using the Game of Checkers,” IBM Journal of
Research and Development, Vol. 3, Issue. 3, 1959.
10
symbolic systems, human experts weren’t needed to predefine steps or game
strategy. Instead, the machine developed intelligence by reviewing data to
determine patterns and then codifying these patterns to inform game strategy.
Validation of Samuel’s work came in 1961 when his checkers program claimed
victory in a live match played against a professional and competitively ranked
human player.
While Samuel made significant progress in machine learning research and model
design during his time at IBM, it wasn’t until after his retirement in 1966 that the
full scope of his findings spread to the broader artificial intelligence community.
According to the authors of Human + Machine: Reimagining Work in the Age of AI,
Paul R. Daugherty and H. James Wilson, news of Samuel’s work was partly inhibited
by his modest character and reluctance to self-promote.4
Over the next two decades, attention took a backseat as other fields of artificial
intelligence including symbolic systems and expert systems5 took precedence in
industry and academic funding. Artificial intelligence, itself, underwent two periods
of declined interest, better known as the two AI winters.
The dot-com era in the 1990s eventually revived investment in machine learning
as a solution to maximize the value of data collected from online retail and digital
systems. While it seems trivial now, access to a large and cheap supply of data to
facilitate learning was a major constraint for AI researchers before this period.
Drastic advances in computer storage and processing capacity also provided the
infrastructure desperately needed to cross the chasm between theory and practical
application.
A new supply of data and cheap computing power handed machine learning a
decisive victory over expert systems as it became more efficient to derive
knowledge from data rather than task experts to configure code as an elaborate
series of if/else rules. Machine learning also offered a comparative advantage in
tackling complex and unknown problems where known steps of reasoning and
action weren’t available, such as detecting fraud and classifying spam email
messages.
4
Paul R. Daugherty and H. James Wilson, “Human + Machine: Reimagining Work in the Age of AI,”
Harvard Business Review Press, 2018.
5
Systems that enabled machines to perform rudimentary reasoning using if/else rules as an alternative
to strict predetermined code.
11
the selling price of other houses in the neighborhood as the dependent variable to
design a prediction model. The prediction model can then predict the value (y) of
a house with an unknown selling price by inputting its features (X) into the
prediction model.
12
subgroups. Unsupervised learning is also useful for taking complex unlabeled data
with a high number of variables and transforming that data into a low number of
synthesized variables that are plottable on a 2-D or 3-D plot as output. Although
the input data has been transformed, the goal is to preserve as much of the data’s
original structure as possible, allowing you to better understand the structure of
the data and identify unsuspected patterns.6
Other popular unsupervised learning tasks include anomaly detection such as
fraudulent transactions or catching manufacturing defects, and automatically
removing outliers and complexity from a dataset before feeding the data to a
supervised learning algorithm.
Reinforcement learning is the third and most advanced category of machine
learning and is generally used for performing a sequence of decisions, such as
playing chess or driving an automobile.
Reinforcement learning is the opposite of unsupervised learning as the output (y)
is known but the inputs (X) are unknown. The output can be considered as the
intended goal (i.e. win a game of chess) and the optimal input is found using a
brute force technique based on trial and error. Random input data is fed to the
model and graded according to its relationship to the target output. In the case of
self-driving vehicles, movements to avoid a crash are graded positively, and in the
case of chess, moves to avoid defeat are rewarded. Over time, the model leverages
this feedback to progressively improve its choice of input variables to achieve its
desired output goal.
The AI company Wayve has released a live video recording of a car learning to
drive, which demonstrates the random and iterative nature of reinforcement
learning. Using sensors and a safety driver who intervenes when the car drifts off-
course, the car learns to navigate the circuit within just 20 minutes of training.
6
Aurélien Géron, “Hands-On Machine Learning with Scikit-Learn and TensorFlow: Concepts, Tools, and
Techniques to Build Intelligent Systems,” O’Reilly Media, 2017.
13
QUIZ
1) A model that predicts the height of adult students based on the height
of their adult relatives is an example of:
a. Supervised learning
b. Unsupervised learning
c. Reinforcement learning
d. Classification
a. Supervised learning
b. Unsupervised learning
d. Reinforcement learning
a. Supervised learning
b. Unsupervised learning
c. Reinforcement learning
a. Supervised learning
b. Unsupervised learning
c. Reinforcement learning
14
a. Distance to city
b. Year built
c. Suburb
d. Price of house
15
SOLUTIONS
1) a, Supervised learning
2) d, Reinforcement learning
3) b, Unsupervised learning
4) a, Supervised learning
5) d, Price of house
16
2
DEVELOPMENT ENVIRONMENT
As the practical exercises delivered in this book use Jupyter Notebook as the
development environment for Python 3, this chapter serves as an optional guide
for installing Jupyter Notebook. If you have prior experience using Jupyter
Notebook or have read my earlier title Machine Learning for Absolute Beginners,
then you may wish to proceed to the next chapter.
Jupyter Notebook is a popular choice for practitioners and online courses alike, as
it combines live code, explanatory notes, and visualizations into one convenient
workspace and runs from any web browser.
Jupyter Notebook can be installed using the Anaconda Distribution or Python’s
package manager, pip. As an experienced Python user, you may wish to install
Jupyter Notebook via pip, and there are instructions available on the Jupyter
Notebook website (http://jupyter.org/install.html) outlining this option. For
beginners, I recommend choosing the Anaconda Distribution option, which offers
an easy click-and-drag setup (https://www.anaconda.com/products/individual/).
This installation option will direct you to the Anaconda website. From there, you
can select your preferred installation for Windows, macOS, or Linux. Again, you
can find instructions available on the Anaconda website based on your operating
system.
After installing Anaconda on your machine, you’ll have access to a number of data
science applications including rstudio, Jupyter Notebook, and graphviz for data
visualization through the Anaconda application. Next, you need to select Jupyter
Notebook by clicking on “Launch” inside the Jupyter Notebook tab.
17
Figure 2: The Anaconda Navigator portal
To initiate Jupyter Notebook, run the following command from the Terminal (for
Mac/Linux) or Command Prompt (for Windows):
jupyter notebook
Terminal/Command Prompt then generates a URL for you to copy and paste it into
your web browser.
Copy and paste the generated URL into your browser to access Jupyter Notebook.
Once you have Jupyter Notebook open in your browser, click on “New” in the top
right-hand corner of the web application to create a new notebook project, and
select “Python 3.”
18
You’ve now successfully set up a sandbox environment in your web browser using
Jupyter Notebook. This means that the following experimentation and code
changes will not affect resources outside of the isolated testing environment.
19
3
Data scientists rarely work alone. This means it’s vital to maintain consistent code
that can be read and reused by other programmers. Similar to using WordPress
plugins with websites, code libraries make it easy for data scientists to perform
common tasks using pre-written modules of code.
With WordPress, for example, you can install a comments management plugin
called Discuz on a portfolio of websites. Using the same plugin for each website
eliminates the need for developers to familiarize themselves with each site’s
underlying code. They simply need to familiarize themselves with the basic
interface and customization settings of the Discuz plugin.
The same logic and benefits apply to machine learning libraries, as complex
algorithms and other functions can be called through the same code interface.
Moreover, rather than writing the statistical requirements of a regression algorithm
over many lines of code, you can call the algorithm from a library such as Scikit-
learn using just one line of code.
Example:
my_model = LinearRegression()
Pandas
Pandas is a library for managing and presenting your data. The name “Pandas”
comes from the term “panel data,” which refers to Panda’s ability to create a series
of panels, similar to sheets in Excel.
Pandas can be used to organize structured data as a dataframe, which is a two-
dimensional data structure (tabular dataset) with labeled rows and columns, similar
to a spreadsheet or SQL table. You can also use Pandas to import and manipulate
an external dataset including CSV files as a dataframe without affecting the source
file as modifications take place inside your development environment.
20
Figure 6: Example of a Pandas’ dataframe
NumPy
NumPy is often used in combination with Pandas and is short for “numeric Python.”
On its own, NumPy is used for managing multi-dimensional arrays and matrices,
merging and slicing datasets, and offers a collection of mathematical functions
including min, max, mean, standard deviation, and variance.
NumPy consumes less memory and is said to perform better than Pandas with
50,000 rows or less.7 NumPy, though, is often used in conjunction with Pandas as
the latter is more user-friendly and easier to interpret in an interactive environment
such as Jupyter Notebook. A Pandas dataframe is also more suitable for managing
a mix of data types, whereas a NumPy array is designed for dealing with numerical
data, especially multi-dimensional data.8
Most machine learning models demonstrated to beginners in massive open online
courses and textbooks structure data as a Pandas dataframe rather than a NumPy
array but often draw on the NumPy library for mathematical and other
miscellaneous operations.
Scikit-learn
Scikit-learn is the core library for general machine learning. It offers an extensive
repository of shallow algorithms9 including logistic regression, decision trees, linear
regression, gradient boosting, etc., a broad range of evaluation metrics such as
mean absolute error, as well as data partition methods including split validation
and cross validation.
Scikit-learn is also used to perform a number of important machine learning tasks
including training the model and using the trained model to predict the test data.
The following table is a brief overview of common terms and functions used in
machine learning from Scikit-learn.
7
Goutham Balaraman, “NumPy Vs Pandas Performance Comparison,” gouthamanbalaraman.com,
March 14, 2017, http://gouthamanbalaraman.com/blog/numpy-vs-pandas-comparison.html
8
Dimensions are the number of variables characterizing the data, such as the city of residence, country
of residence, age, and sex of a user. Up to four variables can be plotted on a scatterplot but three-
dimensional and two-dimensional plots are easier for human eyes to interpret.
9
Shallow algorithms can be roughly characterized as non-deep learning approaches that aren’t
structured as part of a sophisticated network. In shallow learning, the model predicts outcomes directly
from the input features, whereas in deep learning, the output is based on the output of preceding layers
in the model and not directly from the input features.
21
Table 1: Overview of key Scikit-learn terms and functions
Matplotlib
Matplotlib is a visualization library you can use to generate scatterplots, histograms,
pie charts, bar charts, error charts, and other visual charts with just a few lines of
code. While Matplotlib offers detailed manual control over line styles, font
properties, colors, axes, and properties, the default visual presentation is not as
striking and professional as other visualization libraries and is generally used in
conjunction with Seaborn themes.
Seaborn
Seaborn is a popular Python visualization library based on Matplotlib. This library
comes with numerous built-in themes for visualization and complex visual
techniques including color visualization of dependent and independent variables,
sophisticated heatmaps, cluster maps, and pairplots. The combination of
Seaborn’s pre-formatted visual design and Matplotlib’s customizability make it
easy to generate publication-quality visualizations.
Other popular visualization libraries include Plotly (an interactive visualization
Python library) and Cufflinks (which connects Plotly directly
with Pandas dataframes to create graphs and charts).
TensorFlow
A round-up of popular machine learning libraries wouldn’t be complete without an
introduction to Google’s TensorFlow. While Scikit-learn offers a broad set of popular
22
shallow algorithms, TensorFlow is the library of choice for deep learning and
artificial neural networks (ANN).
TensorFlow was created at Google and supports various advanced distributed
numerical computation techniques. By distributing computations on a network with
up to thousands of GPU instances, TensorFlow supports advanced algorithms
including neural networks that would be impossible to run on a single server.
Unfortunately for Mac users, TensorFlow is only compatible with the Nvidia GPU
card, which is no longer available with Mac OS X. Mac users can still run TensorFlow
on their CPU but will need to run their workload on the cloud to access the GPU.
23
QUIZ
1) Which library is better for managing a mix of data types (numeric and
non-numerical data)?
a. NumPy
b. Pandas
c. Seaborn
d. Matplotlib
a. Scikit-learn
b. Pandas
c. Tensorflow
d. Seaborn
a. Scikit-learn
b. Cuffles-learn
c. Matplotlib
d. Seaborn
c. A Scikit-learn function
24
SOLUTIONS
1) b, Pandas
2) a, Scikit-learn
5) True
25
Exploring the Variety of Random
Documents with Different Content
far from low-priced. They are never very exclusive organizations and
yet they give to the strain of the workaday New Yorker his last
lingering trace of hospitality—the hospitality that has lingered around
Bowling Green and Trinity and St. Paul's church-yards since colonial
days and the coffee houses.
* * * * *
"Bless me," he says, "It is going on three o'clock. I've got that
railroad crowd due in my office in fifteen minutes."
That is your dismissal. For ninety minutes he has given you his
hospitality—his rare and unselfish self. He has put the perplexing
details of his business out of his mind and given himself to whatever
flow of talk might suit your fancy. Now the hour and a half of grace
is over—and you are dismissed, courteously—but none the less
dismissed. With your host you descend to the crowded noisome
street. He sees you to the subway—gives you a fine warm grasp of
his strong hand—and plunges back into the great and grinding
machine of business.
Lunch in your Day of Days within the City of the Towers is over.
Three o'clock. Before the last echoes of Trinity's bell go ringing down
through Wall street to halt the busy Exchange—the multitude has
been fed. Miss Stenographer has had her salad and éclair, two
waltzes and perhaps a "turkey trot" into the bargain, and is back at
the keys of her typewriter. Mr. President has entertained that Certain
Party at the club and has made him promise to sign that mighty
important contract. And the certain Party and Mr. President rode for
half an hour on the mechanical horses in the gymnasium. What fun,
too, for those old boys?
Three o'clock! The cashiers are totaling their receipts, the
waiters and the 'buses are upturning chairs and tables to make way
for the scrub-women, some are already beginning to don their
overcoats to go uptown; but the three-quarters of a million of
hungry mouths have been fed. New York has caught its breath in
mid-day relaxation and once more is hard at work—putting in the
last of its hours of the business day with renewed and feverish
energy.
IV
You had planned at first to walk up Broadway. You wanted to see
once again the church-yards around Trinity and St. Paul's, perhaps
make a side excursion down toward Fraunces' Tavern—just now
come back into its own again. Some of the old landmarks that are
still hidden around downtown New York seemed to appeal to you.
But your host at luncheon laughed at you.
"If you want to spend your time that way, all right," he said, "but
the only really old things you will find in New York are the faces of
the young men. You can find those anywhere in the town."
Of course you did. The new Grand Central, with its marvelous
blue ceiling capping a waiting-room so large that the New York City
Hall, cupola, wings and all could be set within it, can hardly escape
the attention of any traveler who passes within its portals.
"That is typical of our restless, lovely city," she tells you. And
you, yourself, have heard that only two years ago they tore down a
nineteen story building at Wall and Nassau streets so that they
might replace it by another of the towers—this one thirty stories in
height.
* * * * *
"You can get more value for less money and less value for much
money in New York than in any other large city in the world," says
Katherine.
She is right—and you know that she is right. You can have a
glorious ride up the street, that even in its days of social decadence
is still the finest highway in the land—a ride that continues across
the town and up its parked rim for long miles—for a mere ten cents
of Uncle Sam's currency and as for the reverse—well you are going
to dinner in a smart hotel with Katherine in a little while.
The great shops dominate the avenue. And if you look with
sharp eyes as the green bus bears you up this via sacre, you may
see that one of the greatest ones—a huge department store encased
in architecturally superb white marble—bears no sign or token of its
ownership or trade. An oversight, you think. Not a bit of it. Four
blocks farther up the avenue is another great store in white marble
—a jewelry shop of international reputation. You will have to scan its
broad façade closely indeed before you find the name of the firm in
tiny letters upon the face of its clock. Oversight? Not a bit of it. It is
the ultra of shop-keeping in New York—the assumption that the shop
is so well known that it need not be placarded to the vulgar world.
And if strangers from other points fail to identify it—well that is
because of their lack of knowledge and the shopkeeper may secretly
rejoice.
But, after all, it is the little shops that mark the character of Fifth
avenue—not its great emporiums. It is the little millinery shops
where an engaging creature in black and white simpers toward you
and calls you, if you are of the eternal feminine, "my dear;" the
jewelry shops where the lapidary rises from his lathe and offers a bit
of craftsmanship; the rare galleries that run from old masters to
modern etchers; specialty shops, filled top to bottom with toys or
Persian rugs, or women's sweaters, or foreign magazines and books,
that render to Fifth avenue its tremendous cosmopolitanism. These
little shops make for personality. There is something in the personal
contact between the proprietor and the customer that makes mere
barter possess a real fascination. And if you do pay two or three
times the real value in the little shop you have just so much more
fun out of the shopping. And there are times when real treasures
may come out of their stores.
* * * * *
You follow her glance to a very simple brick house, upon the
corner of an inconsequential side street. Beside it on Fifth avenue is
an open lot—of perhaps fifty feet frontage, giving to the avenue but
a plain brown wooden fence.
"I expected you to say that," she laughs. "They have wanted to
build upon that lot—time and time again. But when they approach
the owner he laughs at them and declines to consider any offer. 'My
daughter has a little dog,' he says politely, 'It must have a place for
exercise.' We New Yorkers are an odd lot," she laughs. "You know
that the Goelets kept a cow in the lawn of their big house at
Broadway and Nineteenth street until almost twenty years ago—until
there was not a square foot of grass outside of a park within five
miles. And in New York the man who can do the odd thing
successfully is apt to be applauded. You could not imagine such a
thing in Boston or Baltimore or Philadelphia, could you?"
You admit that your imagination would fall short of such heights
and ask Katherine if you are going up to the far end of the 'bus run
—to that great group of buildings—university, cathedral, hospital,
divinity school—that have been gathered just beyond the
northwestern corner of Central Park.
"No, I think not," she quickly decides, "You know that Columbia
is not to New York as Harvard is to Boston. Harvard dominates
Boston, Columbia is but a peg in the educational system of New
York. The best families here do not bow to its fetich. They are quite
as apt to send their boys to Yale or Princeton—even Harvard."
* * * * *
"Why tea?"
V
Dinner is New York's real function of the day. And dinner in New
York means five million hungry stomachs demanding to be filled. The
New York dinner is as cosmopolitan as the folk who dwell on the
narrow island of Manhattan and the two other islands that press
closely to it. The restaurant and hotel dinners are as cosmopolitan as
the others. Of course, for the sake of brevity, if for no other reason,
you must eliminate the home dinners—and read "home" as quickly
into the cold and heavy great houses of the avenue as into the little
clusters of rooms in crowded East Side tenements where poverty is
never far away and next week's meals a real problem. And
remember, that to dine even in a reasonably complete list of New
York's famous eating places—a new one every night—would take
you more than a year. At the best your vision of them must be
desultory.
Six o'clock sees the New York business army well on its way
toward home—the seething crowds at the Brooklyn bridge terminal
in Park Row, the overloaded subway straining to move its fearful
burden, the ferry and the railroad terminals focal points of great
attractiveness. To make a single instance: take that division of the
army that dwells in Brooklyn. It begins its march dinnerward a little
after four o'clock, becomes a pushing, jostling mob a little later and
shows no sign of abatement until long after six. Within that time the
railroad folk at the Park Row terminal of the old bridge have
received, classified and despatched Brooklynward, more than one
hundred and fifty thousand persons—the population of a city almost
the size of Syracuse. And the famous old bridge is but one of four
direct paths from Manhattan to Brooklyn.
Six o'clock sees restaurants and cafés alight and ready for the
two or three hours of their really brisk traffic of the day. There are
even dinner restaurants downtown, remarkably good places withal
and making especial appeal to those overworked souls who are
forced to stay at the office at night. There are bright lights in
Chinatown where innumerable "Tuxedos" and "Port Arthurs" are
beginning to prepare the chop-suey in immaculate Mongolian
kitchens. But the real restaurant district for the diner-out hardly
begins south of Madison square. There are still a very few old hotels
in Broadway south of that point—a lessening company each year—
one or two in close proximity to Washington square. Two of these
last make a specialty of French cooking—their table d'hôtes are
really famous—and perhaps you may fairly say when you are done
at them that you have eaten at the best restaurants in all New York.
From them Fifth avenue runs a straight course to the newer hotels
far to the north—a silent brilliantly lighted street as night comes
"with the double row of steel-blue electric lamps resembling torch-
bearing monks" one brilliant New York writer has put it. But before
the newest of the new an intermediate era of hotels, the Holland,
the nearby Imperial and the Waldorf-Astoria chief among these. The
Waldorf has been from the day it first opened its doors—more than
twenty years ago—New York's really representative hotel. Newer
hostelries have tried to wrest that honor from it—but in vain. It has
clung jealously to its reputation. The great dinners of the town are
held in its wonderful banqueting halls, the well-known men of New
York are constantly in its corridors. It is club and more than club—it
is a clearing-house for all of the best clubs. It is the focal center for
the hotel life of the town.
All New York that dines out does not make for these great places
or their fellows. There are little restaurants that cast a glamour over
their poor food by thrusting out hints of a magic folk named
Bohemians who dine night after night at their dirty tables. There are
others who with a Persian name seek to allure the ill-informed, some
stout German places giving the substantial cheer of the Fatherland,
beyond them restaurants phrasing themselves in the national dishes
and the cooking of every land in the world, save our own. For a real
American restaurant is hard to find in New York—real American
dishes treats of increasing rarity. A great hotel recently banished
steaks from its bills-of-fare, another has placed the ban on pie; and
as for strawberry short-cake—just ask for strawberry short-cake. The
concoction that the waiter will set before you will leave you
hesitating between tears and laughter—ridicule for the pitiful
attempts of a French cook and tears for your thoughts of the tragedy
that has overwhelmed an American institution. Some day some one
is going to build a hotel with the American idea standing back of it
right in the heart of New York. He is going to have the bravery or
the patriotism to call it the American House or the United States
Hotel or Congress Hall or some other title that means something
quite removed from the aristocratic nomenclature that our modern
generation of tavern-keepers have borrowed from Europe without
the slightest sense of fitness; and to that man shall be given more
than mere riches—the satisfaction that will come to him from having
accomplished a real work.
"So good?" you brashly venture. You had not fancied this
production so successful. He does not even assume to hear your
comment. You decide that you will see this particular play at a later
time. You suggest as much to the indifferent creature behind the
wicket. He replies by telling you that he can only give you tickets for
a Monday or Tuesday three weeks hence—and then nothing ahead
of the seventeenth row. Can he not do better than that? He cannot.
He is positive that he cannot. And his positiveness is Gibraltarian in
its immobility. A faint sign of irritation covers his bland face. He
wants you to see that you are taking too much of his time.
"You know they abolished the speculators two years ago," she
explains.
You move on to the little shop with the inviting sign. The
gentleman behind its counters has manners at least. He greets you
with the smile of the professional shopkeeper.
You are not coldly commercial but, despite that smile, merely
apprehensive. And you are beginning to understand New York.
"Four dollars."
Not so bad at that—just the box-office price. You bring out four
greasy one-dollar bills. His eyes fixed upon them, he places a ticket
down upon the counter.
"Do you think that they are four dollars a dozen?" he sallies.
You give him a ten dollar bill this time. You do not kick. Even
though the show is perfectly rotten and the usherette charges you
ten cents for a poorly printed program and scowls because you take
the change from her itching palm, do not complain. You would not
complain even if you knew that the man in the chair next to you
paid only the regular prices, because he happened to belong to the
same lodge as the cousin of the treasurer of the theater, while the
man in the chair next to Katherine paid nothing at all for his seat—
having a relative who advertises in the theater programs. You do not
kick. Complaint has long since been eliminated from the New York
code and you have begun to realize that.
* * * * *
Beer and rarebits, indeed. Sam Blythe tells of the little group of
four who went into a hotel grillroom not far from Forty-second street
and Broadway, who mildly asked for beer and rabbits.
"We asked for beer and rabbits," insisted the host of the little
group. He really did not know his New York.
* * * * *
VII
It has all quite dazed you. You turn toward Katherine as you ride
home with her in the taxicab—space forbids a description of the
horrors and the indignities of the taxicab trust.
"Every night of the year," she replies. "And typical New Yorkers
like it."
We think that you cannot exclude the man who through some
stroke of fortune has accumulated money in a smaller city, and who
has come to New York to live and to spend it. There are many
thousands of him dwelling upon the island of Manhattan; with his
families they make a considerable community by itself. They are
good spenders, good New Yorkers in that they never complain while
the strings of their purses are never tightly tied. They live in smart
apartments uptown, at tremendously high rentals, keep at least one
car in service at all seasons of the year, dine luxuriously in luxurious
eating-places, attend the opera once a week or a fortnight, see the
new plays, keep abreast of the showy side of New York. They are
typical New Yorkers. In an apartment a little further down the street
—which rents at half the figure and comes dangerously near being
called a flat—is another family. This family also attends the new
plays, although it is far more apt to go a floor or even two aloft,
than to meet the speculator's prices for orchestra seats. It also goes
to the opera, and the young woman of the house is in deadly
earnest when she says that she does not mind standing through the
four or five long acts of a Wagnerian matinee, because the nice
young ushers let you sit on the floor in the intermissions. But this
family goes farther than the drama—spoken or sung. It is conversant
with the new books and the new pictures. That same young woman
swings the Phi Beta Kappa key of the most difficult institution of
learning on this continent. And she knows more about the trend of
modern art than half of the artists themselves. And yet she "goes to
business"—is the capable secretary of a very capable man
downtown.
These are typical New Yorkers. So are a family over in the next
block—theirs is frankly a flat in every sense of that despised word.
They have not been in the theater in a dozen years, never in one of
the big modern restaurants or hotels. Yet the head of that family is a
man whose name is known and spoken reverently through little
homes all the way across America. He is a worker in the church,
although not a clergyman, a militant friend of education, although
not an educator, and he believes that New York is the most
thoughtful and benevolent city in the world. And if you attempt to
argue with him, he will prove easily and smilingly, that he is right
and you—are just mistaken. He and his know their New York—a New
York of high Christian force and precept—and they, too, are New
Yorkers.
Are the sane folk right when they say that he does not exist? We
do not think so. We think that Katherine in all her flippancy was
right. They are all typical New Yorkers who sojourn, no matter for
how little a time, within her boundaries. We will go farther still. You
might almost say that all Americans are typical New Yorkers. For
New York is, in no small sense, America. Other towns and cities may
publicly scoff her, down in their hearts they slavishly imitate her, her
store fronts, her fashions, her hotel and her theater customs, her
policemen, even her white-winged street cleaners. They publicly
laugh at her—down in their hearts they secretly adore her.
3
Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.
ebookname.com