Data Science Archives

Pandas and SQL Similarity 3

This is the 3 post in the series of Pandas and SQL command line similarity, link to first and second post are

Finding unique values from the datasets

Pandas

We can find unique value and its count for a column with the help of below command, this may be required for in-depth data analysis.

SQL

Adding column to existing datasets

Sometimes we have to add new column to complete the analysis, procedure to add new column in Pandas and SQL are

Pandas –

This command will add empty column to the existing data frame with data type Object.

SQL

Finding Mean,SUM,Min and Max values of column

The statistical values play important role in the data analysis to develop Machine learning model, the commands to find them are

Pandas

SQL

Drop the column/s

We have to drop some column/s during data analysis which may be irrelevant in developing Machine learning models. The command from Pandas and SQL are

Pandas

SQL

Hope you enjoyed this 3 article to find out Pandas and SQL command line similarity, Stay tuned for the final article. You can find same article on Medium.com here.

Pandas and SQL similarity 2

This article is the next in the series about the topic of Pandas and SQL similarity which we discussed earlier

Pandas and SQL similarity (2)

Check the dtypes

Pandas –

data type in panda — Data types in pandas

To check data types of the data frame, type ‘df.dtypes’ command

SQL –

type sp_help table_name in the query editor. You may be wondering what is ‘nvarchar’

The key difference between varchar and nvarchar is the way they are stored, varchar is stored as regular 8-bit data(1 byte per character) and nvarchar stores data at 2 bytes per character. Due to this reason, nvarchar can hold upto 4000 characters and it takes double the space as SQL varchar.

You can visit for more information on varchar and nvarchar.

Check the data

Pandas –

Once you load the data in to the pandas you can view the data by running the ‘df.head()’ command. This command will give you output of first 5 rows and all columns by default. This output is not in the order.

You can specify the number of rows you want to see e.g. 10

To view last rows in the data frame, type ‘df.tail()’ command

We can sort the data frame by column name in the ascending order

pandas sort values — pandas sort function to sort value by column

by descending order

SQL –

In SQL we have to use ‘ORDER BY’ Column to view the data.

SQL Sort — SQL sorting by ‘ORDER BY’ Clause in Ascending order

SQL Sort1 — SQL sorting by ‘ORDER BY’ Clause in Descending order

Check the NULL Value

Every data set will have some kind of NULL or missing values either in one column or multiple column. Command to find out NULL Value is

Pandas –

df.isnull().sum()

The ‘Product Base Margin’ column has 72 missing values.

pandas null value — pandas isnull function

SQL –

In SQL to get NULL Value we have to run below query for each column.

SELECT * from table where columnname IS NULL

Select the columns

What if we want to see the specific columns, how do we select them?

Pandas –

Note – Pandas will not display all rows in the output by default, to see all rows please run the command –> ‘pd.set_option(‘display.max_rows’, None)’

SQL –

select [Row ID],[Order Priority] from table name.

SQL Select column — SQL select multiple column from the table

Filter the data

Every data scientist or data analyst have to filter the data to get some insight. How do we do that in both pandas and SQL?

Pandas –

Lets say, we need details of all the orders shipped through ‘Regular Air’ only

and sent to California state.

pandas filter command — pandas advance filter function

SQL –

SELECT * from table where [Ship Mode] = ‘Regular Air’

SQL Filter command — SQL advance filter command

Hope you like the article about pandas and SQL command similarity part 2, Stay tuned for next update.

Pandas and SQL Similarity 1

Pandas and SQL are the integral part of Data Science. It is used for data manipulation and analysis while SQL or Structured Query Language is usually used to fetch data, update the contents of the table, or operate on the structure of the database or tables, using any type of database tools, which will have a user interface to apply the operations on the database.

I believe Pandas and SQL are both similar in terms of analyzing the datasets, freshers like me find it difficult to remember the command syntax of both tool. A small effort from my end to show the command similarity.

Let’s Start..

Pandas Installation –

To work on both the tool, first we need to install them. I am using Anaconda distribution on my laptop which comes with all the required library pre-installed. To download Anaconda, please go to their site here

Launch the Jupyter Notebook

Load the pandas module,

Note – If you are not using Anaconda, then pandas installation procedure can be found here

SQL Installation –

I am using Microsoft SQL Express edition for the demo purpose, which is available here for the download and installation procedure here.

Note – The only expectation here is that you should know pandas syntax.

Importing the data –

Pandas –

I am using ‘read_excel’ because my source file is Excel and added sheet_name parameter because i want to import only first sheet from the excel file.

SQL –

You can import entire excel in to the SQL database with the help of SQL Import and Export wizard.

You have to choose source data source as ‘Microsoft Excel’ as shown in below screenshot

Choose the Excel file which you want to import and the correct Excel version

SQL Import Export select destination — SQL Import Export destination

Select ‘SQL Server Native Client 11.0’ in the ‘Choose a Destination’ screen

Pick correct server name, authentication type and the database

You can select first option to copy entire data from the excel sheet or you can write custom SQL query to import only required data. I will go ahead and select first option.