SQL Bridge Tool to Exploratory Data Analysis
edaSQL is a library to link SQL to Exploratory Data Analysis and further more in the Data Engineering. This will solve many limitations in the SQL studios available in the market. Use the SQL Query language to get your Table Results.
Installation
Install dependency Packages before installing edaSQL
pip install pyodbc
pip install ipython
Optional dependency for better visualization - Jupyter Notebook
pip install notebook
Now Install using pip . Offical Python Package Here!!
pip install edaSQL
(OR)
Clone this Repository. Run this from the root directory to install
python setup.py install
Documentation
Read the detailed documentation in readthedocs.io (still under the development)
License
The license for edaSQL is MIT license
Need help?
Stuck on your edaSQL code or problem? Any other questions? Don't hestitate to send me an email ([email protected]).
edaSQL Jupyter NoteBook Tutorial
Access the sample Jupyter Notebook here!!
Access the Sample Data Used in this Repo
- CSV
- DataBase Backup ( you can restore the DB in SQL Studio )
edaSQL for DataFrame: If you are using the CSV or Excel as a source , Read using the Pandas & start from the 3. Data Overview
Import Packages
import edaSQL
import pandas as pd
1. Connect to the DataBase
edasql = edaSQL.SQL()
edasql.connectToDataBase(server='your server name',
database='your database',
user='username',
password='password',
sqlDriver='ODBC Driver 17 for SQL Server')
2. Query Data
sampleQuery = "select * from INX"
data = pd.read_sql(sampleQuery, edasql.dbConnection)
3. Data Overview
insights = edaSQL.EDA(dataFrame=data,HTMLDisplay=True)
dataInsights =insights.dataInsights()
deepInsights = insights.deepInsights()
4. Correlation
eda = edaSQL.EDA(dataFrame=data)
eda.pearsonCorrelation()
eda.spearmanCorrelation()
eda.kendallCorrelation()
5. Missing Values
eda.missingValuesPlot(plot ='matrix')
eda.missingValuesPlot(plot ='bar')
eda.missingValuesPlot(plot ='heatmap')
eda.missingValuesPlot(plot ='dendrogram')
6. Outliers
eda.outliersVisualization(plot = 'box')
eda.outliersVisualization(plot = 'scatter')
outliers = eda.getOutliers()