Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/ngambip/diabetes_factors_2024
Exploring BMI Categories and Health Factors.
https://github.com/ngambip/diabetes_factors_2024
dashboards data datacleaning dax-languague powerbi sql sqlstudio tsql visualization
Last synced: about 1 month ago
JSON representation
Exploring BMI Categories and Health Factors.
- Host: GitHub
- URL: https://github.com/ngambip/diabetes_factors_2024
- Owner: ngambip
- Created: 2024-07-12T09:30:00.000Z (5 months ago)
- Default Branch: main
- Last Pushed: 2024-08-11T14:11:31.000Z (4 months ago)
- Last Synced: 2024-08-12T12:55:57.191Z (4 months ago)
- Topics: dashboards, data, datacleaning, dax-languague, powerbi, sql, sqlstudio, tsql, visualization
- Homepage: https://ngambip.github.io/Diabetes_factors_2024/
- Size: 6.39 MB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# SQL TO POWER BI.
---------------------------------------------------------------------------------------------------------------------------------
## Body Mass Index A numerical value of a person's weight in relation to their height.
------------------------------------------------------------------------------------------------------------------------------------
# Table of contents
- [Objective](#objective)
- [User story](#Userstory)
- [Data Source](#data-source)
- [Stages](#stages)
- [Design](#design)
- [Tools](#tools)
- [Development](#development)
- [Pseudocode](#pseudocode)
- [Data Exploration](#data-exploration)
- [Data Cleaning](#data-cleaning)
- [Testing](#testing)
- [Data Quality Tests](#data-quality-tests)
- [Visualization](#visualization)
- [Results](#results)
- [DAX Measures](#dax-measures)
- [Analysis](#analysis)
- [Findings](#findings)
- [Discovery](#discovery)
- [Recommendations](#recommendations)
- [Conclusion](#conclusion)
- [Project Evaluation](#project-evaluation)
# Objective
Objective of Project
Exploring the correlation between BMI and variables like age, socioeconomic status, diet, and education to identify populations at risk of hypertension and diabetes.
Key Pain Point
Understanding how BMI correlates with various factors to pinpoint vulnerable populations and inform targeted health interventions.
## User Story
I aimed to design a dashboard providing instant analytics on BMI, age, socioeconomic status, education, and diet to identify and support populations at risk of hypertension and diabetes.
The Dashboarrd will provide insights about the potential triggers of diabetes with people of a high BMI measure
- BMI and Age
- BMI and Socio-Economic Factors
- BMI Against Education
- BMI Against Diet
- BMI Against Gender
- BMI Against Ethnicity
I will use this information to to help the implementers of the project to design specific solutions for target groups.
# Data Source
This dataset contains comprehensive health data for 1,879 patients, including critical information such as Patient ID, Demographic Details, and Lifestyle Factors,it offers insights into factors impacting the prevalence of diabetes.[The data is sourced from Kaggle: ](https://www.kaggle.com/datasets/rabieelkharoua/diabetes-health-dataset-analysis)
# Stages
- Design
- Developement
- Testing
- Analysis
# Design
## Dashboard components required
- What should the dashboard contain based on the requirements provided?
The dashbord should answer the following question
1. Which age group has the highest BMI Measure?
2. Does Ethicinity have a factor on BMI Measure?
3. Which educational level has the worst BMI?
4. What is the Average BMI across the Population in Question?
5. Is Gender a factor interms of Average BMI?
6. To what Extent is Diet a Factor to Average BMI?
## Tools Used
Excel: Data exploration
SQL Server: Data cleaning, testing, and analysis
Power BI: Data visualization through interactive dashboards
GitHub: Hosting project documentation
# Development
## Pseudocode
- General Approach
1. Import data into Excel from kaggle
2. Data exploration
3. Upload data into SQL
4. Data cleaning in SQL
5. Data Testing in SQL
6. Upload cleaned data into Power BI
7. Create visuals and analze findings
9. Create repository on Github
## Data Exploration
This is the stage where you have a scan of what's in the data, errors and inconcsistencies.
- Initial Observations:
1. The dataset is voluminous; focused on relevant data matching project objectives.
2. Identified inconsistencies in decimal places, standardized across relevant columns.
3. Categorised all data that was represented by Binary coding .
## Data Cleaning
- Column and Row Checks:
1. Ensured data integrity through column count,
2. row count
3. duplication checks.
4. Verified data types for accuracy.
5. Cheched for Null and Blanks
Working Dataset:
Reduced columns from 46 to 8, focusing on key variables.
Standardized BMI and diet quality score to 4 decimal places for consistency.
Converted all binary coding data to descritive data as guided by the data source.
## Data Testing
Below are the Data Quality Check
## Row count
```sql
/*
# Count the total number of records (or rows) in the SQL view
*/
SELECT
COUNT(*) AS no_of_rows
FROM
Diabetes_factors_2024_View;
```
### Output
## Column count
```sql
/*
# Count the total number of columns (or fields) are in the SQL view
*/
SELECT
COUNT(*) AS column_count
FROM
INFORMATION_SCHEMA.COLUMNS
WHERE_
TABLE_NAME = 'Diabetes_factors_2024_view'
```
### Output
## Data Type Check
```sql
/*
# Check the data types of each column from the view by checking the INFORMATION_SCHEMA view
*/
-- 1.
SELECT
COLUMN_NAME,
DATA_TYPE
FROM
INFORMATION_SCHEMA.COLUMNS
WHERE
TABLE_NAME = 'Diabetes_factors_2024_view';
```
## Duplicate count check
### SQL query
```sql
/*
# 1. Check for duplicate rows in the view
# 2. Group by the PatientId
# 3. Filter for groups with more than one row
*/
-- 1.
SELECT
PatientId
FROM
Diabetes_factors_View
-- 2.
GROUP BY
PatientId
-- 3.
HAVING
COUNT(*) > 1;
```
### Output
# Visualization
This shows the different factors that affect BMI measures and to what extent.
## DAX Measures
### 1. Average BMI Categories
```sql
SWITCH(
TRUE(),
Diabetes_factors_2024_view[BMI] < 18.5, "Underweight",
Diabetes_factors_2024_view[BMI] >= 18.5 && Diabetes_factors_2024_view[BMI] < 24.9, "Normal weight",
Diabetes_factors_2024_view[BMI] >= 25 && Diabetes_factors_2024_view[BMI] < 29.9, "Overweight",
Diabetes_factors_2024_view[BMI] >= 30, "Obese",
"Unknown"
)
```
### Output
Return Average BMI of All Patients
Reference for the above DAX calulation is based on World Health Organization measuring Chart.
### SOCIAL ECONOMIC STATUS
Return Social Economic status againts average BMI
### EDUCATIONAL LEVEL
return Educational Level Against Average BMI
### DIET SCORES
return Average Diet Score Against Average BMI
# Data Analysis
(i) Analyzing the distribution of age and average BMI reveals insights into health demographics concerning BMI categories.
(ii) Examining BMI and socioeconomic factors involves analyzing the distribution of BMI based on various socioeconomic factors to identify correlations and impacts on BMI outcomes.
(iii) Investigating BMI in relation to educational attainment assesses the impact of education on BMI and overall health.
(iv) Exploring BMI categories in connection with diet quality and habits helps understand the dietary influences on BMI
# Findings
(i) BMI Age Analysis: Highlight obesity trends, emphasizing age-specific interventions. Age groups 40-49 and 80-89 had the highest obesity numbers, with all groups having over 100 obese patients.
(ii) BMI Education Analysis: Identified higher obesity rates among individuals with Bachelor's degrees. This group constitutes at least 39% of the population and has a significant number of individuals with average BMI above the recommended levels.
(iii) BMI Socioeconomic Status Analysis: Noted variations in BMI across different socioeconomic groups, with medium-income earners having the highest number of obese patients, recording 302.
(iv) BMI Diet Score Analysis: Individuals with poor and fair diet scores had above-average BMIs. Out of a population of 1,879, 321 had poor diet scores, and 245 had fair diet scores. These groups, categorized as obese, constituted 30% of the population.
(V) BMI Ethnicity Analysis: It was noted that individuals of Caucasian descent had above-average BMIs. Out of the population, 280 Caucasians were recorded as obese and overweight.
(Vi) BMI Gender Analysis: Gender has a minimal impact on average BMI measurements, with only a 1% difference between men and women. Women have a slightly higher average BMI than men.
(Vii) It was also noted that 34% of the entire population was categorized as obese, and 27% was found to be overweight. This means over half of the population is classified as unhealthy and at potential risk for serious diseases.
# Recommendations
(i) Develop targeted interventions to prevent obesity. Focus on early intervention for ages 20-29 to promote healthy lifestyles. Tailor programs for other age groups to address their unique needs and promote overall health.
(ii) BMI by Socioeconomic Status: Tailor health programs to SES. For lower SES, offer subsidized nutrition education and activity programs. For higher SES, focus on wellness programs and health screenings.
(iii) Tailor health campaigns. For lower education, offer simple tips and community programs. For higher education, provide detailed information and specialized wellness activities.
(iV) Tailor health campaigns: offer simple tips and community programs for lower education, and provide detailed information and specialized wellness activities for higher education levels.
# Conclusion
The BMI analysis project has provided valuable insights into the factors influencing obesity across different demographics. The findings underscore the importance of age, education, socioeconomic status, and diet in shaping BMI trends. By implementing targeted interventions and tailored health programs, we can address the specific needs of various groups and promote healthier lifestyles.
# Project Evaluation
This project was a worthwhile undertaking as it highlighted critical areas for intervention and provided actionable recommendations to combat obesity. The insights gained from this analysis can inform public health policies and programs, ultimately contributing to better health outcomes for diverse populations. The project's success lies in its ability to identify at-risk groups and propose practical solutions, making it a valuable contribution to public health research and practice.
