Convert Column Type Transform

Use Convert Column Type to assign a specific data type to one or more columns. This can correct misidentified types and prepare values for analysis.

Basic Usage

To convert column types:

1. Select the Convert Column Type transform from the transform menu.
2. For each column you want to change: a. Choose the field under Columns. b. Select the new data type from the Convert To dropdown.
3. Apply the transformation.

Configuration Options

Basic Options

• Columns: Choose a column to convert. You can add more columns and set a different type for each one.
• Convert To: Choose the new type for each selected column:
  • Numeric
  • Text
  • DateTime
  • Categorical

tip

Add another Columns and Convert To pair for each additional column.

Data Types

Numeric

Numbers used in calculations, including integers and floating-point values.

Best for: Quantities, measurements, and counts.

Text

Character strings, including values that contain numbers but should not be used in calculations.

Best for: Names, descriptions, and codes.

DateTime

Dates and times that you want to sort, calculate with, or analyze over time.

Best for: Dates, timestamps, and time-series data.

Categorical

A limited set of distinct values. Categorical data can use less memory and supports category-specific analysis.

Best for: Classifications, groups, and yes/no values.

Examples

Example: Adjusting Data Types in a Sales Dataset

Input Dataset:

date_column	season_column	sales_column
October 10, 2023	Fall	123.45
10/31/2023	Fall	234.56
November 15, 2023	Fall	345.67
12/31/2023	Winter	678.90

Initial Data Types:

• date_column: object (string)
• season_column: object (string)
• sales_column: object (string)

Configuration:

• date_column: DateTime
• season_column: Categorical
• sales_column: Numeric

Result:

date_column	season_column	sales_column
2023-10-10	Fall	123.45
2023-10-31	Fall	234.56
2023-11-15	Fall	345.67
2023-12-31	Winter	678.90

New Data Types:

• date_column: datetime64[ns]
• season_column: category
• sales_column: float64

The transform converts the dates to datetime values, the seasons to categories, and the sales values to numbers.

Best Practices

1. Confirm what each column represents before changing its type.
2. Watch for information loss when converting to a less precise type, such as float to integer.
3. Check the result for errors or unexpected NaN values.
4. Consider a categorical type for repeated values in large datasets; it can improve performance.
5. Use a consistent date format before converting a column to DateTime.

Troubleshooting

• Numeric conversion fails: Check for non-numeric characters or inconsistent formatting.
• DateTime conversion fails: Make sure every date uses a parseable format. Clean inconsistent values first if needed.
• Too many categories are created: Check whether the column is truly categorical or whether some values should be grouped.