Enjoying the book!

In the “Translating SQL WHERE clauses“ There’s More” section the negation causes Pandas to return salaries < $80000 and >$120000 so it doesn’t emulate a SQL ‘between’.

crime_sort.resample('QS-MAR')['IS_CRIME', 'IS_TRAFFIC'].sum().head() Page 412 describes 'QS_MAR as quarters starting in March, while it should read "ending" in March.
It's important because the resultant table starts at the beginning of Dec, which is not possible under the latter scenario and caused some confusion for me.

When I try to download the file by clicking the download button it just opens another webpage with unformatted csv text information. How do I actually download the file?

pycharm produces a different series than the one in the example. The column keys are lost this way

The code, as written, does not produce the output shown in Out [49] in ipython in a jupyter notebook.

In the scatter plot, color= expects one input, not a list of two inputs.
In the line plot, the x axis has no labels.

I solved this problem by trying different expressions for paths to the chinook.db file for the create_engine(), which eventually produced a useful error message. The following line works if you provide an absolute address for the dir that holds the chinook.db file in the string that I called "path_to_file":

engine = create_engine('sqlite:////' + path_to_file + 'chinook.db')

For absolute addresses, create_engine expects //// not ///.

this was my original statement of the problem

from sqlalchemy import create_engine
engine = create_engine('sqlite:///chinook.db')
tracks = pd.read_sql_table('tracks', engine)
tracks.head()

InvalidRequestError Traceback (most recent call last)
/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/pandas/io/sql.py in read_sql_table(table_name, con, schema, index_col, coerce_float, parse_dates, columns, chunksize)
239 try:
--> 240 meta.reflect(only=[table_name], views=True)
241 except sqlalchemy.exc.InvalidRequestError:

/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/sqlalchemy/sql/schema.py in reflect(self, bind, schema, views, only, extend_existing, autoload_replace, **dialect_kwargs)
4148 "Could not reflect: requested table(s) not available "
-> 4149 "in %r%s: (%s)" % (bind.engine, s, ", ".join(missing))
4150 )

InvalidRequestError: Could not reflect: requested table(s) not available in Engine(sqlite:///chinook.db): (tracks)

During handling of the above exception, another exception occurred:

ValueError Traceback (most recent call last)
in ()
----> 1 tracks = pd.read_sql_table('tracks', engine)
2 tracks.head()

/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/pandas/io/sql.py in read_sql_table(table_name, con, schema, index_col, coerce_float, parse_dates, columns, chunksize)
240 meta.reflect(only=[table_name], views=True)
241 except sqlalchemy.exc.InvalidRequestError:
--> 242 raise ValueError("Table %s not found" % table_name)
243
244 pandas_sql = SQLDatabase(con, meta=meta)

ValueError: Table tracks not found

Chapter 9, Comparing President Trump's and Obama's approval ratings

Code below does not work any more,

base_url = 'http://www.presidency.ucsb.edu/data/popularity.php?pres={}'
trump_url = base_url.format(45)

df_list = pd.read_html(trump_url)
len(df_list)

resulting in ValueError: No tables found

Can you share the data in csv format or modify the source code to work with the present website?

In Ch. 5 - EDA, there are examples with the vehicles.csv dataset, but I cannot seem to locate it in the dataset. Is there a url that can be provided to the actual data? I couldn't locate this on fueleconomy.gov

thanks

.dtypes.value_counts() is the one to go for;
this following is deprecated: .get_dtype_counts()