Unit 3-5 15 M merged

Unit 3
1. Implement the data analysis for multiple data source using multiple connection.

Step 1: Connect to Multiple Data Sources

1. Open Tableau:
o Launch Tableau Desktop.
2. Connect to Data:
o On the start page, click on "Connect".
o Choose the type of data source you want to connect to (e.g., Excel, SQL Server,
Google Sheets, etc.).
o Follow the prompts to establish a connection to your first data source.
3. Add Additional Data Sources:
o Once the first connection is established, go to the Data menu and select "New
Data Source".
o Repeat the process to connect to additional data sources.
o Tableau allows you to connect to multiple types of data sources simultaneously.
Step 2: Join or Blend Data
Depending on how your data is structured, you can either join or blend the data.
Option A: Joining Data Sources
• If both data sources share a common field (e.g., Customer ID, Date), you can join them:
1. Go to the Data Source tab.
2. Drag one data source onto the other to create a join.
3. Select the join type (Inner, Left, Right, Full Outer) and define the join condition
based on the common field.
Option B: Blending Data Sources
• If the data sources do not share a common field, or if they are at different levels of
granularity, use blending:
1. Add a new worksheet.
2. From the Data pane, drag a dimension from the primary data source to the view.
3. Then, drag a measure from the secondary data source; Tableau will
automatically create a blend using the common dimensions.
4. Ensure that both data sources have related fields for effective blending.

Step 3: Create Calculated Fields

• You can create calculated fields to derive new insights from the combined data.
1. Right-click in the Data pane and select "Create Calculated Field."
2. Write your calculation using fields from both data sources, ensuring you use the
correct syntax for referencing fields.
Step 4: Build Visualizations
1. Drag and Drop Fields:
o Start dragging fields from your combined data sources to the rows and columns
shelves to create visualizations.
o Use filters and marks to refine your visualizations and focus on relevant data.
2. Use Dashboard for Integration:
o Create a dashboard to bring together multiple visualizations from your various
data sources.
o Add interactivity (filters, actions) to help users navigate and explore the data
effectively.
Unit 3-5 15 M merged

Step 5: Publish and Share

1. Publish to Tableau Server or Tableau Online:
o Once your analysis is complete, publish the workbook to share insights with
your team or stakeholders.
o Ensure permissions are set appropriately for data security.
2. Schedule Data Refreshes:
o If using Tableau Server, set up data refresh schedules to ensure the dashboard
reflects the most current data.
Step 6: Monitor and Iterate
• Gather Feedback: After sharing the dashboard, gather feedback from users to identify
areas for improvement.
• Iterate on Analysis: Continuously refine the visualizations and data connections based
on evolving business needs.
Conclusion
By following these steps, you can effectively implement data analysis using multiple data
sources in Tableau, enabling you to create comprehensive insights that drive decision-making.
This approach leverages the strengths of different datasets, providing a richer context for
analysis.

2. For any health care data perform the extraction, transformation and finally visualizing
the output using Tableau.
Unit 3-5 15 M merged

Unit 4
1. You are working with customer data that includes customer IDs, names, ages,
purchase amounts, and satisfaction ratings (on a scale of 1 to 5). The dataset is
large and contains missing values, duplicates, and improper data types. Explain the
process you would follow in R to:
• Create appropriate variables for each data field and assign data.
• Use vectors and factors to handle the satisfaction ratings.
• Store the data in a list for later manipulation.
• Convert the variables to appropriate data classes for analysis (e.g., satisfaction
ratings as factors).
• Clean the dataset by removing duplicates and handling missing data.
• Write a function that takes the cleaned data and returns a summary report
including the average purchase amount and satisfaction rating

1. Create Data Frame

# Load library
library(dplyr)
# Create data frame with sample data
customer_data <- data.frame(
customer_id = c(1, 2, 3, 4, 5),
name = c("Alice", "Bob", "Charlie", "David", "Eve"),
age = c(25, 30, 35, NA, 40),
purchase_amount = c(100, 150, NA, 200, 250),
satisfaction_rating = c(5, 4, 3, 2, 5)
)

2. Convert Satisfaction Ratings to Factors

customer_data$satisfaction_rating <- factor(customer_data$satisfaction_rating,
levels = 1:5,
labels = c("Very Unsatisfied", "Unsatisfied", "Neutral", "Satisfied", "Very Satisfied"))

3. Store Data in a List

customer_list <- list(data = customer_data, description = "Customer Data")

4. Convert Variables to Appropriate Classes

customer_list$data$customer_id <- as.integer(customer_list$data$customer_id)
customer_list$data$age <- as.numeric(customer_list$data$age)
customer_list$data$purchase_amount <- as.numeric(customer_list$data$purchase_amount)

5. Clean Dataset (Remove Duplicates & Handle Missing Data)

customer_list$data <- customer_list$data %>%
distinct() %>%
na.omit()

6. Summary Function
summary_report <- function(data) {
avg_purchase <- mean(data$purchase_amount)
avg_satisfaction <- mean(as.numeric(data$satisfaction_rating))
list(average_purchase_amount = avg_purchase,
average_satisfaction_rating = avg_satisfaction)
Unit 3-5 15 M merged

# Generate and print summary

summary_result <- summary_report(customer_list$data)
print(summary_result)

This streamlined version covers all the essential steps to process the customer data efficiently.

2. Develop a decision support system in R for an e-commerce company that

categorizes customers based on their purchase frequency and total spending into
three groups: "Low Value", "Medium Value", and "High Value". The system should:
• Use a for loop to iterate through a dataset of customer IDs, their number of
purchases, and total spending.
• Categorize each customer based on the following conditions:
• If purchases > 10 and spending > 1000, categorize as "High Value".
• If purchases between 5 and 10 and spending between 500 and 1000, categorize as
"Medium Value".
• Otherwise, categorize as "Low Value".
• Use if-else statements to make the decisions.
• Store the results in a list and return a summary of how many customers fall into
each category.
• Discuss how this system could be extended to handle more complex decision-
making criteria (e.g., incorporating customer feedback or satisfaction scores).

Here’s how to develop a decision support system in R for categorizing customers based on their
purchase frequency and total spending:
Implementation Steps
1. Create Sample Customer Data: Define a data frame with customer IDs, number of
purchases, and total spending.
2. Categorize Customers: Use a for loop with if-else statements to classify customers
into "Low Value," "Medium Value," or "High Value."
3. Store Results: Save results in a list and summarize the categories.

Sample Code
# Create sample customer data
customer_data <- data.frame(
customer_id = 1:10,
number_of_purchases = c(12, 7, 15, 4, 10, 3, 9, 8, 5, 11),
total_spending = c(1500, 700, 2000, 300, 900, 200, 600, 800, 550, 1200)
)
# Initialize a list to store results
customer_categories <- list()
# Iterate through each customer and categorize
for (i in 1:nrow(customer_data)) {
purchases <- customer_data$number_of_purchases[i]
spending <- customer_data$total_spending[i]
Unit 3-5 15 M merged

if (purchases > 10 && spending > 1000) {

category <- "High Value"
} else if (purchases >= 5 && purchases <= 10 && spending >= 500 && spending <= 1000) {
category <- "Medium Value"
} else {
category <- "Low Value"
}
customer_categories[[as.character(customer_data$customer_id[i])]] <- category
}
# Convert the list to a data frame for summarization
categories_df <- as.data.frame(table(unlist(customer_categories)))
# Rename columns
colnames(categories_df) <- c("Category", "Count")
# Print the summary of customer categories
print(categories_df)
Summary Output
The resulting data frame categories_df will show the number of customers in each category:
Category Count

High Value X

Medium Value Y

Low Value Z
Extension Ideas for Complex Decision-Making
To enhance the decision support system for more complex decision-making, consider the
following extensions:
1. Incorporating Customer Feedback:
o Gather feedback ratings from customers and include this as a factor in the
categorization process. For instance, customers with high spending but low
satisfaction ratings might be categorized differently.
2. Utilizing Demographic Data:
o Include demographic information (e.g., age, location) to refine customer
categorization and personalize marketing strategies.
3. Seasonal Trends Analysis:
o Implement seasonal analysis to adjust categories based on purchase behavior
changes during holidays or sales periods.
4. Lifetime Value Prediction:
o Use predictive modeling to estimate customer lifetime value based on historical
data, allowing for proactive engagement with high-potential customers.
5. Machine Learning Integration:
o Develop machine learning models to continuously learn and adjust
categorization based on evolving customer behaviors and preferences.
6. Real-Time Analytics:
o Incorporate real-time data processing to adapt customer categorizations
dynamically as new data comes in.
Unit 3-5 15 M merged

These extensions can make the decision support system more robust, enabling the company to
better understand and respond to customer needs, ultimately enhancing customer satisfaction
and loyalty.
Unit 5
1. Power BI is designed to handle various types of data sources and large-scale
reporting needs. Critically analyze the components of Power BI architecture (Power
BI Desktop, Power BI Service, Gateways, Dataflow, etc.). How do these components
work together to enable real-time reporting and data visualization in a cloud-based
environment? Discuss potential challenges in scaling Power BI for large enterprises
and how these can be addressed through the architecture.
Power BI is a powerful business analytics tool that provides interactive visualizations and
business intelligence capabilities with an interface simple enough for end users to create their
reports and dashboards. Its architecture comprises several key components that work together
to facilitate data reporting and visualization, especially in a cloud-based environment. Here's a
critical analysis of the components and their interplay, along with challenges and solutions
related to scaling Power BI for large enterprises.

Power BI Architecture Components

1. Power BI Desktop:
o Function: A Windows application for creating reports and data models. Users
can transform and visualize data, using DAX for calculations.
o Integration: Reports are published to the Power BI Service for sharing.
2. Power BI Service:
o Function: A cloud-based platform for hosting reports and dashboards, enabling
collaboration and real-time data refreshes.
o Integration: Central hub for consuming reports and managing scheduled data
updates.
3. Gateways:
o Function: Software that securely connects on-premises data sources to the
Power BI Service.
o Integration: Essential for accessing real-time data from on-premises databases.
4. Dataflows:
o Function: Tools for data ingestion, transformation, and preparation in the Power
BI Service, leveraging Power Query.
o Integration: Provide a reusable data preparation layer for datasets.
5. Datasets:
o Function: Data models stored in the Power BI Service, serving as the basis for
reports and dashboards.
o Integration: Allow multiple reports to use consistent data.

Real-Time Reporting
These components work together to enable real-time reporting:
• Data Collection: Power BI Desktop gathers data, which is published to the Power BI
Service.
• Real-Time Updates: Gateways ensure data is refreshed in real time for current insights.
• Collaboration: Teams share and work on reports collectively, promoting a data-driven
culture.
Unit 3-5 15 M merged

Challenges in Scaling Power BI

1. Data Volume and Performance:
o Solution: Use aggregation and incremental data refresh strategies to optimize
performance.
2. Data Governance:
o Solution: Implement a governance framework with roles, permissions, and row-
level security.
3. Integration with Legacy Systems:
o Solution: Use data gateways and ETL processes to integrate and prepare data.

4. User Training and Adoption:

o Solution: Provide training resources and foster a community of practice for
knowledge sharing.
5. Cost Management:
o Solution: Monitor usage and optimize licenses to manage costs effectively.
Conclusion
Power BI's architecture facilitates effective real-time reporting and visualization, but scaling for
large enterprises presents challenges that can be addressed through careful planning and
governance strategies. This ensures organizations can fully leverage their data for informed
decision-making.
2. In a global enterprise with different departments and user roles, explain how Power
BI’s sharing and collaboration features can be used to manage access and
distribution of reports and dashboards. Discuss the role of apps, workspaces, and
roles in Power BI Service, and how you would handle challenges related to
governance, data security, and report versioning when sharing across teams. What
strategies would you implement to ensure efficient and secure collaboration?

In a global enterprise with diverse departments and user roles, Power BI’s sharing and
collaboration features are essential for managing access and distribution of reports and
dashboards effectively. Here's how these features, particularly through the use of apps,
workspaces, and roles in Power BI Service, can be leveraged, along with strategies to address
governance, data security, and report versioning challenges.

Power BI Sharing and Collaboration Features

1. Workspaces:
o Definition: Collaborative environments for teams to create, share, and manage
reports.
o User Roles: Different roles (Admin, Member, Contributor, Viewer) control access
and editing capabilities.
2. Apps:
o Definition: Packaged collections of dashboards and reports for easy access by
users.
o Functionality: Admins can control permissions, ensuring only the right users
access sensitive information.

Challenges and Strategies

1. Governance:
Unit 3-5 15 M merged

o Strategy: Implement a governance framework with guidelines for data

ownership and compliance. Monitor usage through audit logs.
2. Data Security:
o Row-Level Security (RLS): Restrict data access based on user roles to protect
sensitive information.
o Permissions Management: Regularly review access permissions at workspace
and app levels.
3. Report Versioning:
o Version Control: Use naming conventions for reports to track changes (e.g.,
Report_v1.0).
o Documentation: Maintain a change log for report modifications.
Collaboration Strategies
1. Clear Roles and Responsibilities: Define user roles to clarify responsibilities within
workspaces and apps.
2. Training and Best Practices: Provide training on data security and governance for users.
3. Regular Audits: Conduct periodic reviews of access permissions and usage statistics.
4. Feedback Mechanism: Implement a way for users to report issues or suggest
improvements.
5. Leverage Power BI Apps: Use apps to organize and distribute related reports efficiently.

Conclusion
By effectively utilizing Power BI’s features and implementing strategies for governance, security,
and collaboration, enterprises can manage access and distribution of reports while
safeguarding sensitive data and ensuring report integrity across teams.