Preprocessing and analyzing a dataset found on kaggle.

The dataset contains 4,525 entries and 11 columns, all of which are of type object. Here's a detailed summary:

property_name: Name and type of the property (4,525 unique entries). areaWithType: Type of area measurement (e.g., Super Area, Carpet Area). square_feet: Area of the property in square feet, with values in the format of xxx sqft. transaction: Type of transaction (e.g., Resale, New Property). This column has 4421 non-null entries, indicating some missing values. status: Status of the property (e.g., Ready to Move). It has 4524 non-null entries. floor: Information about the floor, with entries like 5 out of 10. This column has some missing values as well. furnishing: Furnishing status of the property (e.g., Unfurnished, Semi-Furnished). Contains some missing values. facing: Direction the property faces (e.g., East, West). Missing values are present here. description: Descriptions of the properties, which are highly variable and text-heavy. price_per_sqft: Price per square foot, with values in the format of ₹xxxx per sqft. This column also contains missing values. price: Total price of the property, with various formats including ₹xx Lac and Call for Price.

Missing Values Here are the columns with missing values:

transaction: 104 missing values status: 1 missing value floor: 45 missing values furnishing: 340 missing values facing: 589 missing values description: 1,371 missing values price_per_sqft: 368 missing values Handling Missing Values We will handle missing values using the following strategies:

Drop columns with excessive missing data: For description, we might consider dropping it due to high variability and large number of missing values. Fill with most frequent or default values: For categorical columns like transaction, status, floor, furnishing, and facing. Drop rows with missing values in critical columns: Columns like price and square_feet are critical for analysis.