Errata

The website mentioned in the example requires using a password, which has not yet been discussed in the book. To make this example work, try changing the following:
requests.get("https://cdph.data.ca.gov/api/views/6tej-5zx7/rows.json?accessType=DOWNLOAD")

to this:
requests.get("https://data.ct.gov/resource/y6p2-px98.json?category=Fruit&item=Peaches")

Note from the Author or Editor:
Ch. 5, top p.133 (print edition). Change

response = requests.get(
"https://cdph.data.ca.gov/api/views/6tej-5zx7/rows.json
?accessType=DOWNLOAD")

to

response = requests.get(
"https://chhs.data.ca.gov/api/views/pbxw-hhq8/rows.json?accessType=DOWNLOAD")

$sandpit python -m SimpleHTTPServer

should be changed

$ python -m SimpleHTTPServer

Note from the Author or Editor:
Ch. 2, middle p17 (print edition). Change

Start your development server in the project directory
$sandpit python -m SimpleHTTPServer
Serving HTTP on 0.0.0.0 port 8000 ...

to

Start your development server in the project directory
$ python -m SimpleHTTPServer
Serving HTTP on 0.0.0.0 port 8000 ...

"...very handy enumerate keyword..."

enumerate is a built-in function, not keyword.

Note from the Author or Editor:
ch. 2, middle p.43 (print edition). Just under heading 'Enumerating a List' replace

"... very handy enumerate keyword... "

with

" ... very handy built-in enumerate function ... "

built-in range keyword

should changed to

built-in function range

Note from the Author or Editor:
ch. 2, top p.48. Last bullet point. Change

"... built in range keyword.."

to

"... built in range function..."

It says:
"Adding a comma to the end of the print statement inhibits the
addition of an unnecessary newline."

which means it uses Python 2's print statement.

However, in page 2's sidebar "Python 2 or 3?", the author says he will use Python 3's print function.

Note from the Author or Editor:
ch. 3, top p62, bullet point (1) (print ed.). Change

" ... adding a comma to the end of the print statement ... "

to

"... adding a comma after the print function call inhibits ...

ch.3, p.63.

There are two code samples, the second starting with an 'import csv' line. This line should be placed at the top of the first code sample instead. i.e. change

...
with open('data/nobel_winners.csv...
...
import csv

with open('data/nobel_winners.csv...

to
...
import csv

with open('data/nobel_winners.csv...
...
with open('data/nobel_winners.csv...

ch.3, p.68 top code block

The code block starting

In [0]: time_str = '2012...

should be changed to, in asciidoc:

[source, python]
----
In [0]: from datetime import datetime

In [1]: time_str = '2012/01/01 12:32:11'

In [2]: dt = datetime.strptime(time_str, '%Y/%m/%d %H:%M:%S') <1>

In [3]: dt
Out[2]: datetime.datetime(2012, 1, 1, 12, 32, 11)
----

A footnote needs to be ammended in the paragraph beginning

'One library I've found myself using...'

In asciidoc the following change is needed:

One library I've found myself using a fair deal recently is https://dataset.readthedocs.org/en/latest/[Dataset], a module designed to make working with SQL databases a little easier and more Pythonic than existing powerhouses like SQLAlchemyfootnote:[Dataset's official motto being `databases for lazy people'. Dataset is not part of the standard Anaconda package so you'll want to install it using pip from the command line: `$ pip install dataset'.]. Dataset tries to provide the same degree of convenience you get when working ...

ch.3, p.82.
The code sample in the middle note needs to be changed from:

oid = bson.ObjectId()
oid.generation_time...

to

import bson
oid = bson.ObjectId()
oid.generation_time...

ch.5, p.144, code-block at bottom of page and preceding paragraph:

The reference to 'follower_ids()' should be 'followers_ids()':

.... The Tweepy method followers_ids (get...

then

[source, python]
----
my_follower_ids = api.followers_ids() <1>

for id in my_followers_ids:
followers = api.followers_ids(id) <2>
# ...

The tag <svg id='char'> is not closed; should be <svg id='char'></svg>

Consequence of this typo:
Javascript script.js would not be executed, and not chart shown.

ch.5, p.152, bottom code-block:

Need to change

get_column_titles(wikitable)

to
get_column_titles(table)

In the text of page 155, the function name is get_nationality:

"
A little get_nationality() function will use the winner links ...
"

In the code of page 156, the function name is get_winner_nationality.

Note from the Author or Editor:
ch. 5, bottom p.155 (print ed.). Change

" ... A little get_nationality ... "

to

" ... little get_winner_nationality ... "

ch. 6, p.183, Example 6-5. The code should be changed from

if href:
request = scrapy.Request('https:' + href[0],\

to

if href:
request = scrapy.Request(href[0],\

ch.7, p.202:

The code should be changed from

a = array([45, 65 ...

to

a = np.array([45, 65...

On page 210, before pandas can be used to create the dataframe, it needs to be imported:

import pandas as pd

On page 211, for me to work the loc command needs be invoced with square brackets:

df.loc["Albert Einstein']

Note from the Author or Editor:
ch. 8, p210 para 3 (print ed.).

Change

df = pd.read_json('data/nobel_winners.json')

to

import pandas as pd

df = pd.read_json('data/nobel_winners.json')

p. 211, below second para, bullet point (2). Change

In [3] df.loc('Albert Einstein')

to

ln[3] df.loc['Albert Einstein']

it says:
"
For now, let’s take a quick look at Pandas Panel s, which are
containers for multiple DataFrame...
"

But the section of Panel is at the end of the chapter, not following the current section.

Note from the Author or Editor:
ch.8, second para p.214 (print ed.). Change

... For now let's look take a quick look at Pandas Panels, which are containers for multiple data frames before moving on to see how we create data frames from existing data and how to save the results of our data frame manipulations.

to

For now let's see how we create DataFrames from existing data and how to save the results of our data frame manipulations.

The examples assume that pandas has been importet completely into the modules / consoles namespace.

Following the logic of the preceding expamples it should be

df = pd.DataFrame({...})

Note from the Author or Editor:
ch. 8, p.214, (print ed.) section 'Creating and Saving DataFrames'. In code sections, change "df =DataFrame ... " to "df = pd.DataFrame":

df = DataFrame({ ---> df = pd.DataFrame({
...
df = DataFrame.from_dict... ---> df = pd.DataFrame.from_dict...

If the file format for the to_excel method is not provided, I receive an error. The example needs to read

df.to_excel('nobel_winners.xls', sheet_name='WSheet1')

Note from the Author or Editor:
ch. 8, p.219 just after third para from bottom (print ed.). Change

df.to_excel('nobel_winners' ...

to

df.to_excel('nobel_winners.xlsx', ...

In the example for mongo_to_dataframe, the name of the pandas DataFrame constructor needs to be written in CamelCase:

df = pd.DataFrame(list(cursor))

Note from the Author or Editor:
ch.8, middle p.222 (print ed.). The code needs to be changed from

df = pd.dataframe(list(cursor..

to

df = pd.DataFrame(list(cursor...

In contrast to the examples before, these examples assume that the pandas module has been imported completely to the namespace.
To be in line with previous examples, they should read:

s = pd.Series([1,2,3,4])

etc.

Note from the Author or Editor:
ch.8, p.223-4 (print ed.). The 'Series' calls in the code samples should be prefixed with a 'pd' for six changes in total (three on p.223, three on p.224):

s = Series([... ---> s = pd.Series([...

The command to output the item1 dataframe from the panel should be

pn['item1'] instead of panel['item1']

Note from the Author or Editor:
ch. 8, just after second para p.226 (print ed.). Change

panel['item1']

to

pn['item1']

In the example to replace empty strings, the numpy module is used to access the NaN values with np.nan. However the relevant import statement is not shown.

import numpy as np

Note from the Author or Editor:
ch.9, top p.240 (print ed.). Change

bi_col.replace('', np.nan...

to

import numpy as np
bi_col.replace('', np.nan...

ch.9, p.254, first code-block:

In asciidoc, need to add a comma and forward-slash before the second bullet point (<2>):

[source, python]
----
for i,row in df.iterrows():
try:
pd.to_datetime(row.date_of_death, errors='raise') <1>
except:
print '%s(%s, %d)'%(row.date_of_death.ljust(30),\ <2>
row['name'], i) <3>
----

ch.10, p.286

The code sample following '... plots of tip size against total bill.' should be changed from

g = sns.FacetGrid(tips, col="sex")

to

g = sns.FacetGrid(tips, col="smoker")

ch.10, p.287

The code sample needs changing. Line 5 should be changed from

g.map(plt.regplot, "total_bill"...

to

g.map(sns.regplot, "total_bill"...

ch.11, pages 306-7
References to df_nat on this page and following page 307 should be replaced by df_countries and ref. to 'sort' replaced by 'sort_values':

df_countries = df_nat.set_index('name') -> df_countries = df_countries.set_index(...

change
df_nat.sort('nobel_per_capita'...

to
df.countries.sort_values(by='nobel_wins_per_capita', ascending=False)\

Top of page 307:

change

df_nat[df_nat.nobel_wins > 2]
.sort(...

to

df_countries[df_countries.nobel_wins > 2]\
.sort_values(by='nobel_wins_per_capita', ascending=False)\
.nobel_wins_per_capita.plot(kind='bar')

Chapter 7. Array Indexing and Slicing

a = np.array(arange(8))

=>

a = np.arange(8)

Note from the Author or Editor:
ch.7, middle p.202 (print ed.). Change

a = np.array(arange(8))

to

a = np.arange(8)