qrt — Quant Research Tools
Getting Started
API Reference
Roadmap
q.model — Models
Roadmap
Getting Started
q.bt — Backtesting
Roadmap
q.data — Data
Sample datasets
Data schemas
Local files
Market data sources
Roadmap
q.indicator — Indicators
QRT indicators
Simple moving average (SMA)
Exponential moving average (EMA)
Mean absolute deviation (MADEV)
Volume spike ratio
Price spikes
Log returns
Realized variance
Realized quarticity
Bipower variation
Median realized volatility
Minimum realized volatility
Realized volatility by session
Relative strength
Relative-strength moving average
Relative-strength days
Relative-strength phase
RS new high before price
TA-Lib
pandas-ta-classic
Roadmap
q.cross_section — Cross-sectional analysis
Roadmap
q.feature — Feature lifecycle
Roadmap
q.preprocess — Preprocessing
Missing data
Scaling
Outliers
Encoding
Reduction
Selection
Roadmap
q.calendar — Market time
Roadmap
q.signal — Signals
Roadmap
q.model — Models
Roadmap
q.env — Environment
Roadmap
q.plot — Plotting
Classification curves
Return charts
Tearsheet report
Simulation tests
Trade logs
Saving figures
Roadmap
q.portfolio — Portfolio
Roadmap
q.stats — Statistics
Performance & benchmark
Rolling & calendar returns
Factor analytics
Chained API
Roadmap
q.utils — Utilities
Roadmap
Roadmap
On this page
Model utilities
Leakage-safe data splitting
Edit this page
View source
Report an issue
q.model — Models
Roadmap
q.model Roadmap
Only open work is listed here. Completed items are removed.
Model utilities
PyTorch training loop wrappers
Checkpointing and inference helpers
Time-series-aware dataloaders
Leakage-safe data splitting
Walk-forward splits (
q.model.selection.timeseries_split
)
Purged K-fold with embargo (López de Prado)
Combinatorial purged CV
Split visualization (plot which samples are train/test/embargo)
Back to top
q.model — Models
q.env — Environment