Pandasql -The Best Way to Run SQL Queries in Python
Pandas have come a long way on their own, and are considered second to none when it comes to data handling. Still, there are many
SQL power users who consider
SQL queries nothing less than sacred, and swear by them.
For such users and also for those who chase efficiency in coding (I do agree that SQL Queries are more efficient for some operations!), there is some good news. You can use the, as it is, to do data manipulation inside the python environment. That too in Jupyter Notebooks. Not only that, you can query pandas DataFrame directly using only SQL queries or syntax. If it sounds much like a fantasy, tighten your seat belts and join me in this adventure to marry SQL with Pandas. And did I say, You do not need to install or connect any SQL servers 😎
The saviour is python’s library,
As the libraries’ documentation mentions:
pandasqlallows you to query pandas
SQLsyntax. It works similarly to
pandasqlseeks to provide a more familiar way of manipulating and cleaning data for people new to Python or pandas.
You need to install the Python’s Library,
pandasql first. It’s very simple to install. Use any of the below two methods, both use
- Open the terminal and run
pip install -U pandasql
- Open your Jupyter Notebook and in any cell run
!pip install -U pandasql
There is this one function that is used the most from this library. Its the main function
sqldf takes two parameters.
- A SQL query in string format
- A set of session/environment variables (
It becomes tedious to specify
locals(), hence whenever you import the library, run the following helper function along with. This will make things simple going forward.
from pandasql import sqldf
mysql = lambda q: sqldf(q, globals())
There are many variants of SQL in use, and their syntaxes vary a little. Here in
pandasql uses the
SQLite syntax. Most of the standard SQL language
SQLite understands. However, it adds few features of its own while at the same time it does omit some features. Click Here to read the document that attempts to describe what parts of the SQL language SQLite do and do not support.
pandasql automatically detects any pandas DataFrame. You can call them or query them by their name in the same way you would have done with a SQL table.
We are going to use any one of these two basic code samples.
from pandasql import sqldfmysql = lambda q: sqldf(q, globals()) mysql("SQL Query")
from pandasql import sqldfmysql = lambda q: sqldf(q, globals()) query = ''' SQL Query ''' mysql(query)
Import libraries and Data
For this article, we are going to use the data from the
pandasql library itself. Let us import the dependencies and the data.
import pandas as pd from pandasql import sqldf from pandasql import load_meat, load_births # Importing Data # Bring data in Python environment as pandas DataFrame meat = load_meat() births = load_births()
Let us have a look at the Data.
Read Data using SQL Query
We will read the first 5 rows of data, for the meat and births data frames using SQL. The result shall be similar to what we get from using
# specify globals() or locals() using the following helper function mysql = lambda q: sqldf(q, globals()) mysql("SELECT * FROM meat LIMIT 5;")
mysql("SELECT * FROM births LIMIT 5;")
Joining tables is one of the most common tasks being performed by SQL. Understandably so, as the relational databases have data segregated in separate tables. Hence, SQL users are pretty used to using
join() tables in SQL. We can use the power of SQL JOIN here with pandas DataFrame.
query = ''' SELECT m.date, m.beef, m.veal, m.pork, b.births FROM meat AS m INNER JOIN births AS b ON m.date = b.date; ''' mysql(query)
407 rows × 5 columns
GROUP BY using SQL
The data of meat production is month-wise. We want to see the beef production per year. For that we need to
aggregate. We can do this using the SQL
GROUP BY function.
query = '''SELECT strftime('%Y', date) as year , SUM(beef) as beef_total FROM meat GROUP BY year LIMIT 5; ''' mysql(query)
In the above code, we used SQL query to limit the number of rows for the grouped and aggregated table to 5 rows. But the output and the input, both are not SQL tables. They are pandas DataFrames. And this gives us the liberty to use Pandas functions and methods on the same.
Let us do the same operation, and this time the output shall be the first 10 rows. But the SQL query will give a full table and we will use pandas
head() function to get the final output truncated to 10 rows.
query = '''SELECT strftime('%Y', date) as year , SUM(beef) as beef_total FROM meat GROUP BY year; ''' mysql(query).head(10)
UNION ALL to club multiple variables in SQL
We have beef, pork, and veal as meat types, in separate columns. Here we want all the production values in one column and the identifier in another column. We can use
UNION ALL function from SQL to achieve this easily.
#executing union all statements query = """ SELECT date , 'beef' AS meat_type , beef AS value FROM meat UNION ALL SELECT date , 'veal' AS meat_type , veal AS value FROM meat UNION ALL SELECT date , 'pork' AS meat_type , pork AS value FROM meat UNION ALL SELECT date , 'lamb_and_mutton' AS meat_type , lamb_and_mutton AS value FROM meat ORDER BY 1 """ mysql(query).head(10)
Nested Queries of SQL
In SQL, writing queries within another query is commonplace. The same kind of nesting of queries is possible here as well. We will create one table (or say DataFrame) and without assigning it any variable (or name), we will use that to create another table.
# use queries within queries query = """ SELECT m1.date , m1.beef FROM meat m1 WHERE m1.date IN (SELECT date FROM meat WHERE beef >= broilers ORDER BY date) """ mysql(query)
421 rows × 2 columns
In this article, we saw that how easily we can use SQL queries to operate upon the DataFrames. This gives us a unique opportunity. This weapon can be a potent one in any Data Scientist’s arsenal, who knows SQL and Python, both.
They both are powerful languages and have their respective strengths and weaknesses. Using the method shown in this article, or in other words, using the
pandasql library and
sqldf function, we can use the best and most efficient method to manipulate data, well within the python environment and even Jupyter Notebook. This is music to my ears. I hope you enjoyed the
song too 🤓.
In this article, you saw how to use SQL queries inside python. But if you want to connect the two most powerful workhorses of the Data Science world, SQL and Python. This is not the end, but only the first step towards getting the “Best of Both Worlds”.
Now you can start using Python to work upon your data which rests in SQL Databases. In able to connect to your SQL databases, go thru my article How to Access & Use SQL Database with pyodbc in Python. Once you brought it as DataFrame, then all the operations are usual Pandas operations or SQL queries being operated on Pandas DataFrame as you saw in this article.
Apart from the function of SQL shown in this article, many other popular SQL functions are easily implementable in Python. Read 15 Pandas functions to replicate basic SQL Queries in Python for learning how to do that.
The implied learning in this article was, that you can use Python to do things that you thought were only possible using SQL. There may or may not be straight forward solution to things, but if you are inclined to find it, there are enough resources at your disposal to find a way out. You can look at the mix and match the learning from my book, PYTHON MADE EASY – Step by Step Guide to Programming and Data Analysis using Python for Beginners and Intermediate Level.
About the Author: I am Nilabh Nishchhal. I like making seemingly difficult topics easy and write about them. Check out more at https://www.authornilabh.com/. My attempt to make Python easy and Accessible to all is Python Made Easy.
The media shown in this article are not owned by Analytics Vidhya and are used at the Author’s discretion.