Crate cbtop 

cbtop - Compute Block Top

Real-time load testing and hardware monitoring TUI built on the Brick Architecture.

Design Philosophy

• Test-as-Interface: Every component is a falsifiable Brick (PROBAR-SPEC-009)
• presentar-terminal: All widgets and canvas from presentar-terminal (no custom reimplementation)
• Toyota Way: Jidoka, Poka-Yoke, Genchi Genbutsu principles throughout

Architecture

Layer 4: Load Generators  → SimdLoadBrick, CudaLoadBrick, WgpuLoadBrick
Layer 3: Panels           → Overview, CPU, GPU, PCIe, Memory, Thermal
Layer 2: Analyzers        → Throughput, Bottleneck, Thermal
Layer 1: Collectors       → CPU, GPU, PCIe, Memory, Thermal

Widget Source Policy

All widgets and canvas implementations come from presentar-terminal. cbtop does NOT implement its own widgets. If a widget is missing, it MUST be added to presentar-terminal FIRST, then used here.

Re-exports

pub use brick::Brick;

pub use brick::BrickAssertion;

pub use brick::BrickBudget;

pub use brick::BrickVerification;

pub use brick::BrickGrade;

pub use brick::BrickScore;

pub use brick::Scorable;

pub use brick::fnv1a_f32;

pub use brick::BrickProfiler;

pub use brick::DivergenceReport;

pub use brick::KernelTrace;

pub use app::CbtopApp;

pub use config::Config;

pub use error::CbtopError;

pub use headless::Benchmark;

pub use headless::BenchmarkBuilder;

pub use headless::BenchmarkConfig;

pub use headless::BenchmarkResult;

pub use headless::BenchmarkResults;

pub use headless::ComparisonResult;

pub use headless::LatencyStats;

pub use headless::OutputFormat;

pub use headless::RegressionResult;

pub use headless::ScoreInfo;

pub use headless::SystemInfo;

pub use config::ComputeBackend;

pub use config::WorkloadType;

pub use optimize::AnalysisSummary;

pub use optimize::BaselineEntry;

pub use optimize::BaselineReport;

pub use optimize::BottleneckAnalysis;

pub use optimize::BottleneckEntry;

pub use optimize::BottleneckSeverity;

pub use optimize::CpuCapabilities;

pub use optimize::OptimizationSuite;

pub use optimize::OptimizationValidator;

pub use optimize::RegressionDetector;

pub use optimize::RegressionEntry;

pub use optimize::RegressionReport;

pub use optimize::ValidationResult;

pub use optimize::WorkloadConfig;

pub use baseline::BaselineComparison;

pub use baseline::BaselineValidator;

pub use baseline::GpuClass;

pub use baseline::ServerBaseline;

pub use baseline::SingleComparison;

pub use baseline::SmHealth;

pub use baseline::ThroughputGrade;

pub use baseline::ValidationSummary;

pub use baseline::INDUSTRY_BASELINES;

pub use baseline::TGI_BASELINE;

pub use baseline::TRITON_BASELINE;

pub use baseline::VLLM_BASELINE;

pub use quantize::ggml_type_to_format;

pub use quantize::DequantStrategy;

pub use quantize::GgufError;

pub use quantize::GgufHeader;

pub use quantize::GgufLoader;

pub use quantize::GgufResult;

pub use quantize::GgufTensorInfo;

pub use quantize::GgufValue;

pub use quantize::LayerQuantStats;

pub use quantize::QuantFormat;

pub use quantize::QuantStats;

pub use quantize::QuantizedBrick;

pub use quantize::QuantizedWeights;

pub use paged_kv::BlockId;

pub use paged_kv::CacheStats;

pub use paged_kv::EvictionStrategy;

pub use paged_kv::KvBlock;

pub use paged_kv::PagedKvCache;

pub use paged_kv::PagedKvError;

pub use paged_kv::PagedKvResult;

pub use paged_kv::SeqId;

pub use paged_kv::SequenceInfo;

pub use continuous_batcher::BatchSchedule;

pub use continuous_batcher::BatcherStats;

pub use continuous_batcher::ContinuousBatcher;

pub use continuous_batcher::ExponentialMovingAverage;

pub use continuous_batcher::InferenceRequest;

pub use continuous_batcher::Priority;

pub use continuous_batcher::SchedulingPolicy;

pub use continuous_batcher::SequenceGroup;

pub use continuous_batcher::SpeculativeDecoder;

pub use continuous_batcher::SpeculativeOutput;

pub use continuous_batcher::Token;

pub use continuous_batcher::TokenOutput;

pub use ironman::full_validate;

pub use ironman::quick_validate;

pub use ironman::GateCategory;

pub use ironman::GateResult;

pub use ironman::IronmanScorecard;

pub use ironman::IronmanValidator;

pub use ironman::QualityGate;

pub use ironman::IRONMAN_GATES;

pub use grammar::BuiltComputeBlock;

pub use grammar::ByteSize;

pub use grammar::CompositionMode;

pub use grammar::ComputeBlock;

pub use grammar::ComputeBlockBuilder;

pub use grammar::CpuAffinity;

pub use grammar::DataTransform;

pub use grammar::DataType;

pub use grammar::Dimensions;

pub use grammar::ExecutionContext;

pub use grammar::ExecutionPolicy;

pub use grammar::ExecutionResult;

pub use grammar::ExecutionStrategy;

pub use grammar::GpuDevice;

pub use grammar::GrammarError;

pub use grammar::GrammarResult;

pub use grammar::KernelSpec;

pub use grammar::LinearResourceScale;

pub use grammar::LogResourceScale;

pub use grammar::ObservabilityConfig;

pub use grammar::Operation;

pub use grammar::QosLevel;

pub use grammar::QuantScheme;

pub use grammar::ResourceLimits;

pub use grammar::ResourceMapping;

pub use grammar::ResourceScale;

pub use grammar::RetryPolicy;

pub use grammar::ScaleBinding;

pub use grammar::SimdWidth;

pub use grammar::StrategyLayer;

pub use grammar::TensorSpec;

pub use grammar::WorkloadSpec;

pub use adversarial::AdversarialError;

pub use adversarial::AdversarialResult;

pub use adversarial::AdversarialTactic;

pub use adversarial::AdversarialTestSummary;

pub use adversarial::BitFlipInjector;

pub use adversarial::CancellationToken;

pub use adversarial::CheckedArithmetic;

pub use adversarial::ConfigValidator;

pub use adversarial::InputValidator;

pub use adversarial::MonotonicClock;

pub use adversarial::RecoveryHandler;

pub use adversarial::ResourceLimiter;

pub use adversarial::ResourceUsage;

pub use double_blind::AuditEntry;

pub use double_blind::BlackBoxArtifact;

pub use double_blind::FalsificationClaim;

pub use double_blind::FalsificationCriterion;

pub use double_blind::ReleaseDecision;

pub use double_blind::Role;

pub use double_blind::ScorecardComponent;

pub use double_blind::ScorecardV2;

pub use double_blind::SessionState;

pub use double_blind::VerificationAttempt;

pub use double_blind::VerificationReport;

pub use double_blind::VerificationResult;

pub use double_blind::VerificationSession;

pub use tracing_escalation::EscalationReason;

pub use tracing_escalation::EscalationThresholds;

pub use tracing_escalation::OtlpSpanAttributes;

pub use tracing_escalation::SyscallBreakdown;

pub use tracing_escalation::TraceResult;

pub use tracing_escalation::TracingEscalation;

pub use roofline::BatchRooflineAnalysis;

pub use roofline::BatchSummary;

pub use roofline::BottleneckType;

pub use roofline::HardwareProfile;

pub use roofline::RooflineAnalysis;

pub use roofline::RooflinePlot;

pub use roofline::RooflinePlotPoint;

pub use roofline::WorkloadMetrics;

pub use fuzz::bound_value;

pub use fuzz::checked_add_u64;

pub use fuzz::checked_mul_u64;

pub use fuzz::safe_div;

pub use fuzz::sanitize_float;

pub use fuzz::test_float_edge_cases;

pub use fuzz::test_u64_edge_cases;

pub use fuzz::FuzzFailure;

pub use fuzz::FuzzInputValidator;

pub use fuzz::FuzzResult;

pub use fuzz::FuzzSuite;

pub use fuzz::FuzzSummary;

pub use fuzz::FuzzTargetConfig;

pub use fuzz::FuzzValidationError;

pub use statistics::bootstrap_ci;

pub use statistics::percentile;

pub use statistics::trimmed_mean;

pub use statistics::ComparisonResult as StatisticalComparison;

pub use statistics::EffectCategory;

pub use statistics::EffectSize;

pub use statistics::MannWhitneyResult;

pub use statistics::OutlierFilter;

pub use statistics::StatisticalAnalysis;

pub use cache_analysis::elementwise_working_set;

pub use cache_analysis::matrix_working_set;

pub use cache_analysis::optimal_matmul_tile;

pub use cache_analysis::AccessPattern;

pub use cache_analysis::BandwidthPrediction;

pub use cache_analysis::CacheConfig;

pub use cache_analysis::CacheLevel;

pub use cache_analysis::WorkingSetAnalysis;

pub use latency_distribution::DistributionShape;

pub use latency_distribution::HistogramBucket;

pub use latency_distribution::LatencyDistribution;

pub use latency_distribution::LatencyHistogram;

pub use latency_distribution::TailSeverity;

pub use variance_analysis::VarianceAnalysis;

pub use variance_analysis::VarianceInput;

pub use variance_analysis::VarianceSource;

pub use profile_persistence::BackendConfig as ProfileBackend;

pub use profile_persistence::ProfileConfig;

pub use profile_persistence::ProfileError;

pub use profile_persistence::ProfileManager;

pub use profile_persistence::ProfileOverlay;

pub use profile_persistence::ProfileResult;

pub use profile_persistence::WorkloadConfig as ProfileWorkload;

pub use golden_trace::GoldenComparator;

pub use golden_trace::GoldenTrace;

pub use golden_trace::GoldenTraceError;

pub use golden_trace::GoldenTraceManager;

pub use golden_trace::GoldenTraceResult;

pub use golden_trace::SyscallBreakdown as TraceSyscallBreakdown;

pub use golden_trace::SyscallBreakdownDelta;

pub use golden_trace::TraceComparison;

pub use golden_trace::TraceMetrics;

pub use thermal_prediction::analyze_thermal;

pub use thermal_prediction::assess_throttle_risk;

pub use thermal_prediction::CooldownRecommendation;

pub use thermal_prediction::RiskCategory;

pub use thermal_prediction::ThermalAnalyzer;

pub use thermal_prediction::ThermalCorrelation;

pub use thermal_prediction::ThermalPrediction;

pub use thermal_prediction::ThermalSample;

pub use thermal_prediction::ThermalVariance;

pub use thermal_prediction::ThrottleRisk;

pub use thermal_prediction::DEFAULT_THROTTLE_THRESHOLD_C;

pub use thermal_prediction::MIN_SAMPLES_FOR_ANALYSIS;

pub use backend_regression::Backend;

pub use backend_regression::BackendComparison;

pub use backend_regression::BackendMeasurement;

pub use backend_regression::BackendRecommendation;

pub use backend_regression::BackendRegressionDetector;

pub use backend_regression::BackendSummary;

pub use backend_regression::SizeCliff;

pub use backend_regression::TransferAnalysis;

pub use backend_regression::WorkloadType as BackendWorkload;

pub use correlation_analysis::CorrelationAnalyzer;

pub use correlation_analysis::CorrelationResult;

pub use correlation_analysis::EventSample;

pub use correlation_analysis::EventType;

pub use correlation_analysis::InterferenceCategory;

pub use correlation_analysis::InterferenceResult;

pub use correlation_analysis::IsolationAction;

pub use correlation_analysis::IsolationRecommendation;

pub use correlation_analysis::PerformanceSample;

pub use correlation_analysis::SystemSnapshot;

pub use performance_prediction::DataPoint;

pub use performance_prediction::FittedModel;

pub use performance_prediction::ModelType;

pub use performance_prediction::PerformancePredictor;

pub use performance_prediction::Prediction;

pub use performance_prediction::MIN_SAMPLES_FOR_FIT;

pub use anomaly_detection::Anomaly;

pub use anomaly_detection::AnomalyDetector;

pub use anomaly_detection::AnomalyReport;

pub use anomaly_detection::AnomalySeverity;

pub use anomaly_detection::AnomalyType;

pub use anomaly_detection::ChangePoint;

pub use anomaly_detection::DEFAULT_IQR_MULTIPLIER;

pub use anomaly_detection::DEFAULT_ZSCORE_THRESHOLD;

pub use anomaly_detection::MIN_SAMPLES_FOR_DETECTION;

pub use workload_characterization::ClassificationResult;

pub use workload_characterization::RecommendedBackend;

pub use workload_characterization::WorkloadCategory;

pub use workload_characterization::WorkloadCharacterizer;

pub use workload_characterization::WorkloadFeatures;

pub use export_reporting::BenchmarkMetric;

pub use export_reporting::BenchmarkReport;

pub use export_reporting::ComparisonEntry;

pub use export_reporting::ComparisonReport;

pub use export_reporting::ExportFormat;

pub use export_reporting::ReportBuilder;

pub use export_reporting::ReportExporter;

pub use export_reporting::ReportType;

pub use adaptive_threshold::LearnedThreshold;

pub use adaptive_threshold::ThresholdCheck;

pub use adaptive_threshold::ThresholdDirection;

pub use adaptive_threshold::ThresholdLearner;

pub use adaptive_threshold::DEFAULT_CONFIDENCE_LEVEL;

pub use adaptive_threshold::DEFAULT_OUTLIER_THRESHOLD;

pub use adaptive_threshold::MIN_SAMPLES_FOR_LEARNING;

pub use frequency_control::CpuFrequencyInfo;

pub use frequency_control::CpuGovernor;

pub use frequency_control::FrequencyController;

pub use frequency_control::FrequencyLock;

pub use frequency_control::FrequencyReading;

pub use frequency_control::FrequencyVariance;

pub use context_regression::BaselineEntry as ContextBaselineEntry;

pub use context_regression::ContextRegressionPredictor;

pub use context_regression::RegressionCheck;

pub use context_regression::RegressionThreshold;

pub use context_regression::SystemContext;

pub use context_regression::Trend;

pub use context_regression::DEFAULT_COLD_START_MARGIN;

pub use context_regression::DEFAULT_STALENESS_SEC;

pub use context_regression::MIN_SAMPLES_FOR_CONTEXT;

pub use alerting::alert_from_anomaly;

pub use alerting::Alert;

pub use alerting::AlertChannel;

pub use alerting::AlertRouter;

pub use alerting::AlertRouterConfig;

pub use alerting::AlertSeverity;

pub use alerting::DeliveryResult;

pub use alerting::MessageTemplate;

pub use prometheus::validate_metric_name;

pub use prometheus::CounterValue;

pub use prometheus::GaugeValue;

pub use prometheus::HistogramBuckets;

pub use prometheus::HistogramValue;

pub use prometheus::Labels;

pub use prometheus::MetricDef;

pub use prometheus::MetricType;

pub use prometheus::MetricsRegistry;

pub use prometheus::DEFAULT_BUCKETS;

pub use prometheus::DEFAULT_MAX_LABELS;

pub use cost_tracker::default_gpu_pricing;

pub use cost_tracker::BudgetAlert;

pub use cost_tracker::CloudProvider;

pub use cost_tracker::CostComparison;

pub use cost_tracker::CostResult;

pub use cost_tracker::CostTracker;

pub use cost_tracker::EnergyMeasurement;

pub use cost_tracker::GpuPricing;

pub use cost_tracker::DEFAULT_CARBON_INTENSITY;

pub use cost_tracker::JOULES_PER_KWH;

pub use event_streaming::compress_data;

pub use event_streaming::event_from_sample;

pub use event_streaming::EventBatch;

pub use event_streaming::EventStreamer;

pub use event_streaming::MetricEvent;

pub use event_streaming::RetryConfig;

pub use event_streaming::SinkHealth;

pub use event_streaming::SinkType;

pub use event_streaming::DEFAULT_BATCH_SIZE;

pub use event_streaming::SCHEMA_VERSION;

pub use remote_agent::AggregatedResult;

pub use remote_agent::AggregationStrategy;

pub use remote_agent::AuthMethod;

pub use remote_agent::CommandResult;

pub use remote_agent::HostBenchmark;

pub use remote_agent::HostConfig;

pub use remote_agent::HostHealth;

pub use remote_agent::HostState;

pub use remote_agent::RemoteAgent;

pub use remote_agent::RemoteAgentConfig;

pub use remote_agent::RemoteError;

pub use remote_agent::RemoteResult;

pub use remote_agent::DEFAULT_HEALTH_CHECK_INTERVAL_SEC;

pub use remote_agent::DEFAULT_MAX_CONCURRENT;

pub use remote_agent::DEFAULT_RETRY_DELAY_MS;

pub use profile_compare::BenchmarkProfile;

pub use profile_compare::ChangeDirection;

pub use profile_compare::CompareConfig;

pub use profile_compare::CompareError;

pub use profile_compare::CompareResult;

pub use profile_compare::ComparisonVerdict;

pub use profile_compare::EffectMagnitude;

pub use profile_compare::EffectSizeResult;

pub use profile_compare::MetricComparison;

pub use profile_compare::MetricSamples;

pub use profile_compare::ProfileComparator;

pub use profile_compare::ProfileComparison;

pub use profile_compare::WelchTestResult;

pub use profile_compare::DEFAULT_CONFIDENCE_LEVEL as COMPARE_DEFAULT_CONFIDENCE_LEVEL;

pub use profile_compare::DEFAULT_REGRESSION_THRESHOLD as COMPARE_DEFAULT_REGRESSION_THRESHOLD;

pub use profile_compare::MIN_COMPARISON_SAMPLES;

pub use observability_backend::format_dogstatsd;

pub use observability_backend::BackendHealth;

pub use observability_backend::DatadogConfig;

pub use observability_backend::ExportMetric;

pub use observability_backend::ExportResult;

pub use observability_backend::HoneycombConfig;

pub use observability_backend::MetricExportType;

pub use observability_backend::NewRelicConfig;

pub use observability_backend::ObservabilityBackend;

pub use observability_backend::ObservabilityConfig as ObsBackendConfig;

pub use observability_backend::ObservabilityError;

pub use observability_backend::ObservabilityExporter;

pub use observability_backend::ObservabilityResult;

pub use observability_backend::OtlpConfig;

pub use observability_backend::WebhookConfig;

pub use observability_backend::DEFAULT_BATCH_SIZE as OBS_DEFAULT_BATCH_SIZE;

pub use observability_backend::DEFAULT_FLUSH_INTERVAL_MS;

pub use regression_pipeline::BenchmarkMetric as PipelineBenchmarkMetric;

pub use regression_pipeline::BenchmarkResults as PipelineBenchmarkResults;

pub use regression_pipeline::GitRef;

pub use regression_pipeline::MetricRegression;

pub use regression_pipeline::PipelineConfig;

pub use regression_pipeline::PipelineError;

pub use regression_pipeline::PipelineResult;

pub use regression_pipeline::PipelineStatus;

pub use regression_pipeline::RegressionAnalysis;

pub use regression_pipeline::RegressionPipeline;

pub use regression_pipeline::StatusCheck;

pub use regression_pipeline::DEFAULT_REGRESSION_THRESHOLD as PIPELINE_REGRESSION_THRESHOLD;

pub use regression_pipeline::DEFAULT_TIMEOUT_SEC;

pub use regression_pipeline::DEFAULT_WARNING_THRESHOLD;

pub use federated_metrics::AggregatedMetrics;

pub use federated_metrics::FederatedHost;

pub use federated_metrics::FederationConfig;

pub use federated_metrics::GCounter;

pub use federated_metrics::LwwRegister;

pub use federated_metrics::MetricsFederation;

pub use federated_metrics::OrSet;

pub use adaptive_ml::AdaptiveThresholdMl;

pub use adaptive_ml::AnomalyResult as MlAnomalyResult;

pub use adaptive_ml::ClassificationMetrics as MlClassificationMetrics;

pub use adaptive_ml::LearnedWorkloadThreshold;

pub use adaptive_ml::MlThresholdConfig;

pub use adaptive_ml::MlThresholdError;

pub use adaptive_ml::TimeSeriesFeatures;

pub use adaptive_ml::WorkloadClass;

pub use incremental_snapshot::DeltaMetric;

pub use incremental_snapshot::DeltaSnapshot;

pub use incremental_snapshot::IncrementalSnapshotStore;

pub use incremental_snapshot::MetricData;

pub use incremental_snapshot::ProfileSnapshot;

pub use incremental_snapshot::RetentionTier;

pub use incremental_snapshot::SnapshotConfig;

pub use incremental_snapshot::SnapshotError;

pub use incremental_snapshot::SnapshotIndex;

pub use incremental_snapshot::SnapshotQuery;

pub use predictive_scheduler::HostProfile;

pub use predictive_scheduler::InstanceType;

pub use predictive_scheduler::PredictiveScheduler;

pub use predictive_scheduler::PredictiveSchedulerConfig;

pub use predictive_scheduler::SchedulerMetrics;

pub use predictive_scheduler::SchedulingDecision;

pub use predictive_scheduler::WorkloadSpec as SchedulerWorkloadSpec;

Modules

adaptive_ml

Dynamic Adaptive Thresholds with ML (PMAT-049)

adaptive_threshold

Adaptive Threshold Learning System (PMAT-037)

adversarial

Adversarial Falsification Testing (PMAT-019)

alerting

Real-Time Alert Integration System (PMAT-040)

anomaly_detection

Anomaly Detection Engine (PMAT-034)

app

cbtop Application State Machine

backend_regression

Cross-Backend Regression Detector (PMAT-031)

baseline

Industry Baseline Validation (PMAT-016)

brick

Core Brick trait and types (PROBAR-SPEC-009 alignment)

bricks

Brick implementations for cbtop

cache_analysis

Cache Efficiency Analysis Module (PMAT-025)

config

Configuration for cbtop

context_regression

Context-Aware Regression Predictor (PMAT-039)

continuous_batcher

ContinuousBatcher Implementation (PMAT-015)

correlation_analysis

Multi-Metric Correlation Analysis (PMAT-032)

cost_tracker

Cost and Energy Efficiency Tracker (PMAT-042)

double_blind

Double-Blind Verification Framework (PMAT-020)

error

Error types for cbtop

event_streaming

Structured Event Streaming (PMAT-043)

export_reporting

Multi-Format Export System (PMAT-036)

federated_metrics

Federated Metrics Aggregation (PMAT-048)

frequency_control

CPU Frequency Control Backend (PMAT-038)

fuzz

Fuzz Testing Integration (PMAT-023)

golden_trace

Golden Trace Comparison (PMAT-029)

grammar

Grammar of ComputeBlock (§32)

headless

Headless benchmark mode for CI/CD and AI agent integration

incremental_snapshot

Incremental Profile Snapshots (PMAT-050)

ironman

Ironman Falsification Suite (F901-F920)

latency_distribution

Latency Distribution Analysis Module (PMAT-026)

observability_backend

Observability Backend Integrations (PMAT-046)

optimize

Optimization identification tooling for cbtop

paged_kv

PagedKvCache Implementation (PMAT-014)

performance_prediction

Performance Prediction Model (PMAT-033)

predictive_scheduler

PMAT-051: Predictive Scheduling Optimizer

profile_compare

Profile Diffing and A/B Comparison (PMAT-045)

profile_persistence

Profile Persistence and Rotation (PMAT-028)

prometheus

Prometheus Metrics Exporter (PMAT-041)

quantize

QuantizedBrick Implementation (PMAT-013)

regression_pipeline

CI/CD Regression Detection Pipeline (PMAT-047)

remote_agent

Remote SSH/Headless Agent Integration (PMAT-044)

ring_buffer

SIMD-optimized ring buffer for time-series data

roofline

Roofline Model Analyzer (PMAT-022)

statistics

Statistical Analysis Module (PMAT-024)

thermal_prediction

Thermal Trend Prediction (PMAT-030)

tracing_escalation

Tracing Escalation Framework (PMAT-021)

variance_analysis

Variance Source Analysis Module (PMAT-027)

workload_characterization

Workload Characterization System (PMAT-035)

Structs

BrailleGraph

Time-series graph widget.

CellBuffer

Buffer of terminal cells with dirty tracking.

Color

RGBA color with values in the range [0.0, 1.0].

Constraints

Layout constraints that specify minimum and maximum sizes.

DiffRenderer

Differential renderer that minimizes terminal I/O.

DirectTerminalCanvas

Direct terminal canvas that implements presentar’s Canvas trait.

Meter

Horizontal meter widget displaying a percentage value.

Point

A 2D point with x and y coordinates.

Rect

A rectangle defined by position and size.

Size

A 2D size with width and height.

Table

Scrollable table widget with headers and rows.

TextStyle

Text style for rendering.

Enums

ColorMode

Terminal color capability mode.

GraphMode

Render mode for the graph.

Traits

Canvas

Canvas trait for paint operations.