Enum DerivedValue 

#[non_exhaustive]
pub enum DerivedValue {
    Scalar(f64),
}

Output value of a derived metric.

Derived metrics carry an f64 scalar. The f64 carrier is chosen because the value range varies across derivations:

• [0, 1] ratios: cpu_efficiency, affine_success_ratio, involuntary_csw_ratio.
• [0, ∞) ratios: disk_io_fraction (readahead can pull more block-device bytes than the syscall requested, so the ratio exceeds 1.0 in practice).
• [0, ∞) per-event means: avg_wait_ns, avg_slice_ns, avg_iowait_ns — sum over count, both non-negative.
• (-∞, ∞) signed differences: live_heap_estimate = allocated_bytes - deallocated_bytes can go negative when the deallocation total exceeds the allocation total (a freelist drains memory allocated before capture began, or the per-thread TSD counters were sampled mid-update on a thread that has just released a large arena).

All four shapes flow through the same f64 carrier. The per-derivation auto-scale ladder lives on DerivedMetricDef::ladder (not on the value type) so the renderer picks the right magnitude (ns / Bytes / unitless) per row regardless of whether the value is positive, zero, negative, fractional, or in the millions. The is_ratio flag on DerivedMetricDef toggles between the auto-scaled path (e.g. 1.500ms, 7.500GiB) and the raw three-decimal path (0.873 for ratios).

Sign preservation: the auto_scale step uses abs() for the threshold check but propagates the original signed value through the scaled output, and format_derived_value_cell / format_derived_delta_cell both render with {value:.2} or {value:.3} formatters that preserve the explicit - for negatives. The auto_scale_preserves_sign_on_negative_input regression test pins this for the Bytes and ns ladders.

Variants (Non-exhaustive)

Non-exhaustive enums could have additional variants added in future. Therefore, when matching against variants of non-exhaustive enums, an extra wildcard arm must be added to account for any future variants.

Scalar(f64)

Floating-point value. Render via the DerivedMetricDef::ladder + DerivedMetricDef::is_ratio pair: ratios format with three decimals (0.873, +0.100); ladder-bearing values (ScaleLadder::Ns / ScaleLadder::Bytes / etc.) route through the same auto-scale ladders the main table uses.

Implementations

impl DerivedValue

pub fn as_f64(&self) -> f64

Return the underlying f64. Helper for delta math downstream of DerivedRow consumers.

Trait Implementations

impl Clone for DerivedValue

fn clone(&self) -> DerivedValue

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for DerivedValue

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl PartialEq for DerivedValue

fn eq(&self, other: &DerivedValue) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Copy for DerivedValue

impl StructuralPartialEq for DerivedValue