Practical 1

Perform the analysis for the following:

a. Import the data warehouse data in Microsoft Excel and create the Pivot table and Pivot
Chart.
b. Import the cube in Microsoft Excel and create the Pivot table and Pivot Chart to
perform data analysis.

a.
 '

Output:

₹ 1,400.00
₹ 1,200.00
₹ 1,000.00 10th
₹ 800.00 11th
₹ 600.00
12th
₹ 400.00
13th
₹ 200.00
14th
₹ 0.00
India India India India India India 15th

1 2 3 4 5 6
b.
Output:

100000
90000
80000
70000
60000
50000
Dim Customer Count
40000
30000 Dim Time Count
20000
10000
0
1 2 3 4 5 6
Output:
 Practical 2
Apply the what – if Analysis for data visualization. Design and generate necessary reports
based on the data warehouse data. Use Excel.

1. On the Data tab, in the Forecast group, click What-If Analysis.

2. Click on Scenario Manager

The Scenario Manager dialog box appears.

3. Add a scenario by clicking on Add.

4. Type a name (60% highest), select cell C4 (% sold for the highest price) for the Changing
cells and click on OK.
5. Enter the corresponding value 0.6 and click on OK again.

6. Next, add 4 other scenarios (70%, 80%, 90% and 100%).

Finally, your Scenario Manager should be consistent with the picture below:

OR
 Practical 3
Perform the data classification using classification algorithm using R/Python.

Code:
> rainfall <- c(799,1174.8,865.1,1334.6,635.4,918.5,685.5,784.2,985,882.8,1071)
> rainfall.timeseries <- ts(rainfall,start = c(2021,1),frequency = 12)
> print(rainfall.timeseries)
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov
2021 799.0 1174.8 865.1 1334.6 635.4 918.5 685.5 784.2 985.0 882.8 1071.0
> png(file = "rainfall.png")
> plot(rainfall.timeseries)
> dev.off()
null device
1
> plot(rainfall.timeseries)

Output:
 Practical 4
Perform the data clustering using clustering algorithm using R/Python.

Code:
> newiris <- iris
> newiris$Species <- NULL
> (kc <- kmeans(newiris,3))
> table(iris$Species,kc$cluster)
> plot(newiris[c("Sepal.Length","Sepal.Width")],col=kc$cluster)
> points(kc$centers[,c("Sepal.Length","Sepal.Width")],col=1:3,pch=8,cex=2)

Output:
 Practical 5
Perform the Linear regression on the given data warehouse data using R/Python.

Code:
> x <- c(151,174,138,186,128,136,179,163,152,131)
> y <- c(63,81,56,91,47,57,76,72,62,48)
> relation <- lm(y~x)
> png(file = "linearregression.png")
> plot(y,x,col = "blue",main = "Height & Weight Regression",abline(lm(x~y)),cex=1.3, pch=16,
xlab="Weight in Kg", ylab="Height in cm")
> dev.off()

Output:
 Practical 6
Perform the logistic regression on the given data warehouse data using R/Python.

Code:
> quality<-read.csv("C:/quality.csv")
> str(quality)
> table(quality$PoorCare)
> install.packages("caTools")
> library(caTools)
> qualityTrain = subset(quality,split==TRUE)
> qualityTest =subset(quality,split==FALSE)
> nrow(qualityTrain)
> nrow(qualityTest)
> QualityLog=glm(PoorCare~OfficeVisits + Narcotics,data=qualityTrain,family = binomial) >
summary(QualityLog)
> predictTrain=predict(QualityLog,type = "response")
> summary(predictTrain)

> tapply(predictTrain, qualityTrain$PoorCare, mean)

> table(qualityTrain$PoorCare,predictTrain>0.5)
10/25
70/74
> table(qualityTrain$PoorCare,predictTrain>0.7)
8/25
73/74
> table(qualityTrain$PoorCare,predictTrain<0.2)
16/25
54/74

> install.packages("ROCR")
> library(ROCR)
> ROCRpred=prediction(predictTrain,qualityTrain$PoorCare)
> ROCRperf=performance(ROCRpred,"tpr","fpr")
> plot(ROCRperf)
> plot(ROCRperf,colorize=TRUE)
> plot(ROCRperf,colorize=TRUE,print.cutoffs.at=seq(0,1,by=0.1),text.adj=c(-0.2,10.7))
 Practical 7
Write a Python program to read data from a CSV file, perform simple data analysis, and
generate basic insights. (Use Pandas is a Python library).

Code:
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

#Load the CSV file into a DataFrame

file_path = 'C:\WineQT.csv'
df = pd.read_csv(file_path)
#df = pd.read.csv(file_path)

#Summary statistics
summary_stats = df.describe()
print("Summary Statistics")
print(summary_stats)

#Correlation matrix
correlation_matrix = df.corr()
print("\nCorrelation Matrix: ")
print(correlation_matrix)

#Quality distribution
quality_distribution = df['quality'].value_counts().sort_index()
print("\nQuality Distribution: ")
print(quality_distribution)

#Correlation with quality

quality_correlation = correlation_matrix['quality'].sort_values(ascending=False)
print("\nQuality Correlation: ")
print(quality_correlation)

# Correlation with quality

quality_correlation = correlation_matrix ['quality'].sort_values (ascending=False)
print("\nCorrelation with Quality:")
print (quality_correlation)

#Plotting
plt.figure (figsize=(10, 6))
#Quality distribution plot
plt.subplot (2, 2, 1)
sns.countplot (x='quality', data=df, palette='viridis')
plt.title('Quality Distribution')

#Heatmap of correlation matrix

plt.subplot(2, 2, 2)
sns.heatmap (correlation_matrix, annot=True, cmap='coolwarm', fmt='.2f', linewidths=0.5)
plt.title('Correlation Matrix Heatmap')

#Alcohol vs Quality
plt.subplot (2, 2, 3)
sns.boxplot (x='quality', y='alcohol', data=df, palette='viridis')
plt.title('Alcohol vs Quality')

#Density vs Quality
plt.subplot (2, 2, 4)
sns.boxplot (x='quality', y='density', data=df, palette='viridis')
plt.title('Density vs Quality')
plt.tight_layout ()
plt.show()

Output:
 Practical 8
Perform data visualization
a. Perform data visualization using Python on any sales data.

Code:
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# Step 1: Load the CSV data into a pandas DataFrame

def load_data(csv_file):
try:
data = pd.read_csv('C:\sales_data.csv', encoding="ISO-8859-1")
print("Data loaded successfully!")
print(data)
return data
except Exception as e:
print(f"Error loading data: {e}")
return None

# Step 2: Perform basic data inspection

def data_summary(data):
print("\nFirst 5 rows of the data:")
print(data.head()) # Displays the first 5 rows

print("\nData Structure:")
print(data.info()) # Get information on the DataFrame such as column types and non-null
counts

print("\nStatistical summary of numeric columns:")

print(data.describe()) # Gives summary statistics for numeric columns

# Step 3: Visualize the data

# Visualization 1: Total Sales Over Time

def plot_sales_over_time(data):
# Convert 'Date' to datetime if it's not already in the correct format
data['Date'] = pd.to_datetime(data['Date'])

# Aggregate data by Date and calculate the total sales for each day
sales_by_date = data.groupby('Date')['Sales_Amount'].sum().reset_index()
 plt.figure(figsize=(10, 6))
sns.lineplot(x='Date', y='Sales_Amount', data=sales_by_date, marker='o')
plt.title('Total Sales Over Time')
plt.xlabel('Date')
plt.ylabel('Total Sales ($)')
plt.xticks(rotation=45)
plt.tight_layout()
plt.show()

# Visualization 2: Sales Distribution by Region

def plot_sales_by_region(data):
plt.figure(figsize=(10, 6))
sns.boxplot(x='Region', y='Sales_Amount', data=data)
plt.title('Sales Distribution by Region')
plt.xlabel('Region')
plt.ylabel('Sales Amount ($)')
plt.tight_layout()
plt.show()

# Visualization 3: Sales vs Quantity Sold (Scatter Plot)

def plot_sales_vs_quantity(data):
plt.figure(figsize=(10, 6))
sns.scatterplot(x='Quantity_Sold', y='Sales_Amount', data=data, hue='Product',
palette='viridis')
plt.title('Sales Amount vs Quantity Sold')
plt.xlabel('Quantity Sold')
plt.ylabel('Sales Amount ($)')
plt.tight_layout()
plt.show()

# Visualization 4: Top Products by Sales

def plot_top_products_by_sales(data):
top_products = data.groupby('Product')['Sales_Amount'].sum().reset_index()
top_products = top_products.sort_values('Sales_Amount', ascending=False).head(10)

plt.figure(figsize=(10, 6))
sns.barplot(x='Sales_Amount', y='Product', data=top_products, palette='Blues_d')
plt.title('Top 10 Products by Sales Amount')
plt.xlabel('Sales Amount ($)')
plt.ylabel('Product')
plt.tight_layout()
plt.show()
# Main function to load data and generate visualizations
def main():
csv_file = 'sales_data.csv' # Replace with your actual CSV file path
data = load_data(csv_file)

if data is not None:

# Step 2: Data summary and inspection
data_summary(data)

# Step 3: Data Visualizations

plot_sales_over_time(data)
plot_sales_by_region(data)
plot_sales_vs_quantity(data)
plot_top_products_by_sales(data)

# Entry point of the program

if __name__ == "__main__":
main()

Output:
b. Perform data visualization using PowerBI on any sales data.

Step 1: Install and Open Power BI

1. Download & Install Power BI Desktop

o If you haven't installed it yet, download it from Power BI Download.
o Install and open Power BI Desktop.

Step 2: Load the Sales Data

1. Open Power BI Desktop.

2. Click on "Home" > "Get Data" > "Text/CSV".
3. Browse and select the sales_data_sample.csv file you uploaded.
4. Click Open, then click Load (you can check the data preview before loading).

Step 3: Transform Data in Power Query Editor

Before building reports, let's clean and format the data:

1. Click on "Transform Data" to open Power Query Editor.

2. Review columns and check for missing values:

o ADDRESSLINE2, STATE, TERRITORY have missing values, so you can remove them or fill
in with "Unknown".
o To remove a column: Right-click on the column header > Click Remove.
3. Convert ORDERDATE column to Date format:
o Select ORDERDATE column.
o In the top menu, click Transform > Data Type > Date.

4. Ensure SALES, PRICEEACH are in Decimal Number format.

 5. Click "Close & Apply".

Step 4: Create Basic Visualizations

1. Sales Performance Overview

● Total Sales & Orders

1. Go to Report View (bottom left panel).
 2. Drag SALES to a Card Visual (this shows total revenue).
3. Drag ORDERNUMBER to another Card Visual (shows total orders).

● Sales Trend Over Time

1. Drag ORDERDATE to X-Axis in a Line Chart.
2. Drag SALES to Y-Axis (shows sales over time).
2. Sales by Product Line

1. Insert a Bar Chart.

2. Drag PRODUCTLINE to X-Axis.
3. Drag SALES to Y-Axis (shows revenue per product category).

3. Sales by Country

1. Insert a Map Visual (Globe icon).

2. Drag COUNTRY to Location field.
3. Drag SALES to Values field (shows sales by country).
4. Order Status Breakdown

1. Insert a Pie Chart.

2. Drag STATUS to Legend.
3. Drag ORDERNUMBER to Values (shows percentage of orders in each status).

Step 5: Add Filters and Slicers

● Click on "Slicer" and add YEAR_ID to filter by year.

● Add CUSTOMERNAME to filter by customer.
● Add DEALSIZE to filter by deal size.
Step 6: Save & Publish Report

1. Click File > Save As and save the Power BI report.

2. Click Publish to share on Power BI Service (if needed).

Steps to Perform Data Visualization in Power BI

1. Import the Data:

o Open Power BI Desktop.
o Click on "Get Data" > "Excel" and select the uploaded file.
o Load the relevant sheets into Power BI.
2. Data Preparation:
o Check for missing values or inconsistent data.
o Apply transformations (if needed) using Power Query Editor.
3. Create Key Visuals:
o Sales Trends: Use a Line Chart to show sales performance over time.
o Regional Sales Analysis: Use a Map Visual to display sales by region.
o Top Products Sold: Use a Bar Chart to highlight best-selling products.
o Revenue Breakdown: Use a Pie Chart or Treemap to show sales by category.
o Customer Segmentation: Use Clustered Bar Charts to group customers based
on sales.

4. Add Filters and Slicers:

o Include Date Slicers to filter data by month, quarter, or year.
o Use Dropdown Filters for region, product category, or customer segment.
5. Enhance with DAX Measures:
o Create calculated measures such as:
o Total Sales = SUM(Sales[Revenue])
o Sales Growth = ( [Total Sales] - PREVIOUSMONTH([Total Sales]) ) /
PREVIOUSMONTH([Total Sales])
o Use KPIs for profit margin and sales targets.

6. Publish and Share:

o Save the report and publish it to Power BI Service.
o Share with stakeholders via dashboards.
 Practical 9
Create the Data staging area for the selected database using SQL.

1. Load Data into Power BI

Open Power BI Desktop.

Click Home > Get Data > Excel Workbook.
Browse and select sales_data_sample.xlsx.
Click Transform Data to open Power Query Editor instead of directly loading the data.
2. Perform Data Transformations in Power Query
A. Inspect and Rename Columns
Check all column names and rename them to follow a clear naming convention (e.g., Order_Date instead
of order date).

B. Remove Unnecessary Columns

If there are extra columns not needed for analysis, remove them to improve efficiency.

C. Handle Missing Data

Check for null values in key fields such as Order_ID, Customer_Name, Product_Code, etc.
Apply one of the following:
Use Remove Rows > Remove Blank Rows for completely empty records.

Use Replace Values to fill nulls with "Unknown" (for text) or 0 (for numeric fields).
Use Fill Down (for fields where previous values should be carried forward).
D. Change Data Types
Ensure that each column has the correct data type:
Dates: Convert to Date type (Order_Date, Ship_Date).
Numbers: Convert Quantity, Price, Total_Amount to Decimal Number or Whole Number.
Text: Convert Customer_Name, Product_Name, Region to Text.
E. Remove Duplicates
Identify and remove duplicates based on Order_ID or Invoice_ID (whichever is the unique identifier).

F. Standardize Text Formatting

Convert text fields to Proper Case (e.g., Product Name, Customer Name):
Text.Proper([Column_Name])
Trim spaces from text fields to avoid mismatches:
Text.Trim([Column_Name])
3. Create Staging Tables
Fact Table (FactSales):
Keep transactional data such as Order_ID, Date, Product_Code, Quantity, Total_Sales.

Dimension Tables:
DimCustomers: Extract unique Customer_ID, Customer_Name, Region.
DimProducts: Extract unique Product_Code, Product_Name, Category.

DimDates: Create a Date Table if not present (Order_Date, Year, Month, Quarter).
To create a Date Table in Power Query:
Go to New Source > Blank Query.
Open Advanced Editor and paste:
let
StartDate = #date(2020, 1, 1),
EndDate = #date(2030, 12, 31),
DateList = List.Dates(StartDate, Number.From(EndDate - StartDate) + 1, #duration(1, 0, 0, 0)),
DateTable = Table.FromList(DateList, Splitter.SplitByNothing(), {"Date"}),
ChangedType = Table.TransformColumnTypes(DateTable, {{"Date", type date}})
in
ChangedType
Click Close & Apply.
4. Define Relationships in Power BI Model

Go to Model View and establish relationships:

FactSales[Customer_ID] → DimCustomers[Customer_ID]
FactSales[Product_Code] → DimProducts[Product_Code]
FactSales[Order_Date] → DimDates[Date]
5. Create Data Validation Reports

Use DAX Measures to check data quality:

Total Sales:
Total_Sales = SUM(FactSales[Total_Amount])

Count of Missing Customers:

Missing_Customers = COUNTROWS(FILTER(FactSales, ISBLANK(FactSales[Customer_ID])))
Duplicate Orders Check:
Duplicate_Orders = COUNTROWS(FactSales) - DISTINCTCOUNT(FactSales[Order_ID])

Create a Table Visual for missing data analysis.

6. Save & Publish

Click Close & Apply to load the transformed data into Power BI.
 Practical 10
Create the cube with suitable dimension and fact tables based on ROLAP, MOLAP and
HOLAP model.

Step 1.Click File -> New -> Project ->Business Intelligence Projects ->select Analysis Services
Project-> Assign Project Name -> Click OK.

Step 2. In Solution Explorer, Right click on Data Source -> Click New Data Source

Step 3. Click on New

Step 4. Creating New Connection.
Step 5. Click on Test Connection and verify for its success.

Step 6. Select Connection created in Data Connections-> Click Next.

Step 7. Select Option Inherit.

Step 8. Assign Data Source Name -> Click Finish

Step 9. In the Solution Explorer, Right Click on Data Source View -> Click on New Data Source
View.
Step 10. Click Next.

Step 11. Select Relational Data Source we have created previously (Sales_DW)-> Click Next.
Step 12. First move your Fact Table to the right side to include in object list.

Step 13.Select Fact Table in Right Pane (Fact product Sales) -> Click On Add Related Tables.

Step 14. Assign Name (SalesDW DSV)-> Click Finish

Step 15. Now Data Source View is ready to use.

Step 16. In Solution Explorer -> Right Click on Cube-> Click New Cube.
Step 17. Click Next.
Step 18. Select Option Use existing Tables -> Click Next.

Step 19. Select Fact Table Name from Measure Group Tables (FactProductSales) -> Click Next.

Step 20. Choose Measures from the List which you want to place in your Cube --> Click Next.
Step 21. Select All Dimensions here which are associated with your Fact Table-> Click Next

Step 22. Assign Cube Name (SalesDW2) -> Click Finish.

Step 23. Now your Cube is ready, you can see the newly created cube and dimensions added in
your solution explorer.
Step 24.In Solution Explorer, double click on dimension Dim Product -> Drag and Drop Product
Name from Table in Data Source View and Add in Attribute Pane at left side.

Step 25.Double click On Dim Date dimension -> Drag and Drop Fields from Table shown in
Data Source View to Attributes-> Drag and Drop attributes from leftmost pane of attributes to
middle pane of Hierarchy.

Step 26. In Solution Explorer, right click on Project Name (Analysis Services Project3)
Click Properties.
Step 27.In Configuration Properties, Select Deployment-> Assign Your SQL Server Instance
Name Where Analysis Services Is Installed (mubin-pc\fairy) (Machine Name\Instance Name) ->
Choose Deployment Mode Deploy All as of now ->Select Processing Option Do Not Process ->
Click OK.

Step 28. In Solution Explorer, right click on Project Name (AnalysisServicesProject)

Click Deploy.
Step 29. Once Deployment will finish, you can see the message Deployment Completed in
deployment Properties.

Step 30. Open SQL Server Configuration Server→Click on SQL Analysis Server→ Copy the
account name and close the window.

Step 31. In Solution Explorer, right click on Project Name (AnalysisServicesProject3)

→Click Process.

Step 32. Click on Run Button to process the 1stcube.

Step 33. Once processing is complete, you can see Status as Process Succeeded
-->Click Close to close both the open windows for processing one after the other.
Step 34. Now go to SQL Server Management Studio, disconnect the existing database and
connect to Analysis Services→ in the database folder you will see the 1stcube has been
successfully created.