This project involves working with a clinical trial dataset containing information on 500 patients, of which 350 participated in a trial comparing two insulin treatments: Novodra (injectable) and Auralin (oral).
The dataset includes patient details, treatment records, HbA1c measurements, and adverse reactions.

The main goal of data wrangling here is to:

• Clean and organize raw data
• Handle missing or inconsistent values
• Prepare data for analysis (statistical testing, visualization, reporting)

Contains demographic and baseline details:

• Identifiers (patient_id, name, contact, address)
• Demographics (sex, birthdate, age)
• Measurements (weight, height, BMI)

Tracks treatment progress and effectiveness:

• Insulin doses (Auralin, Novodra)
• HbA1c levels (start, end, change)

Logs reported side effects for both treatment groups.

The wrangling process will include:

1. Loading Data – Import CSV/Excel files into Pandas
2. Exploring Structure – Use .info(), .head(), .describe()
3. Cleaning
   • Remove duplicates
   • Standardize column names
   • Handle missing values (imputation or removal)
4. Transformations
   • Convert datatypes (e.g., birthdate → datetime, zip_code → string)
   • Calculate derived columns (e.g., age from birthdate, BMI categories)
5. Merging Tables – Combine patients, treatments, and adverse reactions for complete analysis
6. Validation – Ensure correct ranges (BMI, age ≥ 18, HbA1c values)

This project works with a clinical trial dataset of 500 patients, where 350 participated in a study comparing two insulin treatments: Novodra (injectable) and Auralin (oral).
The dataset includes patient demographics, treatment details, HbA1c levels, and reported adverse reactions.

The goal is to clean, transform, and prepare the data for analysis.

• patient_id → Unique patient ID
• assigned_sex → Sex at birth (Male/Female)
• given_name, surname → Patient names
• address, city, state, zip_code, country → Contact details (all US)
• contact → Phone & email
• birthdate → Patient’s date of birth (Age ≥ 18 included)
• weight, height, bmi → Body stats (Inclusion BMI: 16–38)

• given_name, surname → Patient identifiers
• auralin → Baseline and final insulin doses (units “u”)
• novodra → Same as above, for Novodra group
• hba1c_start, hba1c_end → HbA1c levels at start and end (%)
• hba1c_change → Change in HbA1c (start − end)

• given_name, surname → Patient identifiers
• adverse_reaction → Reported side effect

1. Load Data → Import CSV/Excel files into Pandas
2. Explore → Use .info(), .head(), .describe()
3. Clean → Remove duplicates, standardize names, handle missing values
4. Transform → Convert datatypes, derive new columns (e.g., Age, BMI category)
5. Merge → Combine patients, treatments, and adverse reactions
6. Validate → Ensure correct ranges (Age ≥ 18, BMI 16–38, valid HbA1c values)

df.info() df.describe() df.isna().sum() df.drop_duplicates() df.fillna() df.merge() df.groupby()