Decimal

Fixed-precision decimal values for financial and high-precision computations

Overview

The Decimal type in Internet Object provides a fixed-precision decimal representation designed for applications that require exact numeric values, especially when dealing with financial calculations or other scenarios where floating-point precision issues could lead to significant errors.

Unlike standard floating-point numbers (which may suffer from approximation issues), Decimal types store exact numeric values with a defined precision and scale, ensuring accurate and predictable arithmetic operations.

Representation and Syntax

A Decimal value in Internet Object is represented as a number with the m suffix:

123.45m
-98765.4321m
0.000123m

Basic Decimal Literals

// Simple decimal values
123m         // Integer decimal
123.45m      // Fractional decimal
0.001m       // Leading zeros
-789.01m     // Negative decimal

Scientific Notation

Decimal values also support scientific notation:

1.23e2m      // Equivalent to 123m
1.23e-2m     // Equivalent to 0.0123m
5e3m         // Equivalent to 5000m

Key Concepts

Precision and Scale

Each Decimal value is defined by two key properties:

  • Precision (M): The total number of significant digits in the number

  • Scale (D): The number of digits after the decimal point

For example, in 123.45m:

  • The precision is 5 (total digits: 1,2,3,4,5)

  • The scale is 2 (decimal digits: 4,5)

Fixed-Precision Arithmetic

Decimal values maintain their precision throughout operations, making them ideal for financial calculations where exact values are required:

// With regular floating-point:
0.1 + 0.2 ≈ 0.30000000000000004 // Approximation error

// With decimal type:
0.1m + 0.2m = 0.3m // Exact representation

Rounding Behavior

When precision or scale constraints require rounding:

  1. If a value has more decimal places than the specified scale, it's rounded using the "half up" rounding method (≥ 5 rounds up)

  2. If the rounded value would exceed the precision, a validation error is raised

Schema Definition and Validation

When used with Internet Object schemas, decimal types can be precisely defined and validated:

{
  amount: {
    type: decimal,
    precision: 15,    // Total digits
    scale: 4,         // Decimal places
    min: 0,           // Minimum value constraint
    max: 10000.0000   // Maximum value constraint
  }
}

The decimal type supports various validation properties:

  • precision: Restricts the total number of significant digits

  • scale: Controls the number of decimal places

  • min/max: Validates value range

  • choices: Limits valid values to a predefined set

  • optional: Specifies if the field is optional

  • null: Determines if null values are allowed

Operations and Behavior

Type Conversion

Internet Object provides mechanisms for converting between different numeric types:

// Converting to decimal from other formats
decimal("123.45")    // From string
decimal(123.45)      // From number

Decimal Conversion and Compatibility

Decimal values with different precision/scale configurations can be converted to ensure compatibility during operations:

// Converting between decimal configurations
123.45m.convert(8, 4)    // Converts to precision=8, scale=4 → 123.4500m
123.4567m.convert(5, 2)  // Converts to precision=5, scale=2 → 123.46m (rounds)

The convert method allows:

  • Increasing scale: Adds zeros to the right (e.g., 123.45 → 123.4500)

  • Decreasing scale: Truncates with rounding (e.g., 123.456 → 123.46)

  • Adjusting precision: Ensures the new value fits within specified constraints

When working with decimals of different formats:

  1. Two decimals must have the same precision and scale for direct comparison.

  2. Decimals with different configurations must be converted to a common format.

  3. Conversion may fail if the target precision cannot accommodate the value.

It's important to note that both the integer and fractional parts of a decimal must fit within the precision minus scale. For example:

// This would fail because 12345 requires 5 digits, but precision(4)-scale(2)=2 only allows 2 integer digits
12345.67m.convert(4, 2)  // Error: Value exceeds precision

// This works because the value fits within the precision constraints
123.45m.convert(5, 2)    // 123.45m (no change needed)

Example of decimal comparison with conversion:

let price = 19.95m;        // precision=4, scale=2
let newPrice = 19.95000m;  // precision=7, scale=5

// Cannot compare directly (different precision/scale)
// Must convert to a common format first:
price.convert(7, 5) == newPrice;  // true, both as precision=7, scale=5

Comparisons

Decimal values can be compared with other decimal values as expected:

12.34m > 12.33m    // true
12.34m == 12.34m   // true
12.34m < 12.35m    // true

Use Cases and Examples

Decimal types are particularly valuable in:

  1. Financial applications - currency calculations, banking, accounting

  2. Scientific computing - when exact decimal representation matters

  3. Regulatory compliance - when calculations must be exactly reproducible

  4. Monetary APIs - for consistent data exchange with financial systems

Financial Transaction Example

{
  transactionId: "TX12345",
  amount: 1299.99m,
  currency: "USD",
  tax: 78.00m,
  total: 1377.99m
}

Scientific Measurement Example

{
  experiment: "E-201",
  measurement: 0.00000123m,
  uncertainty: 0.00000002m,
  unit: "meters"
}

Best Practices

  1. Explicitly Define Precision/Scale: Always specify the precision and scale for decimal types in schemas to ensure data consistency.

  2. Use for Financial Data: Prefer decimal type over floating-point for monetary values to avoid rounding errors.

  3. Consider Storage Implications: Decimal types require more storage than standard numbers due to their exact representation.

  4. Range Validation: Use min/max constraints to ensure decimal values stay within expected business bounds.

  5. Manage Precision/Scale When Combining Decimals: When performing operations across decimals with different precision/scale, explicitly convert them to a common format or ensure your target format can accommodate the result.

Technical Considerations

When implementing or working with decimal values, keep the following points in mind:

  1. Precision: Decimal types preserve exact values without rounding errors, unlike floating-point numbers. This makes them ideal for financial and high-precision applications.

  2. Interoperability: Decimal types are compatible with database systems and financial applications that require exact decimal arithmetic, ensuring seamless data exchange and integration.

  3. Performance: While operations on decimal values may be slower than native floating-point operations, they provide guaranteed precision, which is crucial for applications where accuracy is paramount.

  4. Consistency: Decimal calculations produce the same results regardless of the platform or implementation, ensuring reliable and predictable outcomes across different environments.

  5. Implementation Model: The underlying implementation of the Decimal type uses a coefficient-exponent model, similar to database systems like SQL Server and Oracle. This provides a strong basis for interoperability with enterprise data systems and ensures consistent behavior.

PreviousBigIntNextSpecial Numeric Integer Formats

Last updated

Was this helpful?