GAFIME

Verify that GAFIME is correctly installed and all components are working. Use when the user just installed GAFIME, wants to verify their setup, asks "is GAFIME working", "test my installation", "health check", "verify install", "check if GPU is working with GAFIME", or encounters import errors after installation.

Detect the user's hardware platform and recommend optimal GAFIME engine configuration. Use when the user asks about hardware compatibility, which backend to use, GPU detection, VRAM availability, CUDA version, Apple Silicon support, or wants to know the best EngineConfig for their system. Also use when the user says things like "what GPU do I have", "will GAFIME work on my machine", "detect my hardware", or "configure for my system".

Diagnose why a GAFIME native compute backend is not loading. Use when the user reports that GAFIME is slow, GPU is not being used, CUDA/Metal backend failed to load, DLL or shared library loading errors, or says things like "why is GAFIME using CPU", "CUDA not working", "backend not found", "gafime is slow", or "GPU not detected".

Compare GAFIME-discovered features against manually crafted features to show the value of automated feature interaction mining. Use when the user wants to benchmark GAFIME, compare automated vs manual feature engineering, justify using GAFIME, or says things like "is GAFIME better than manual features", "compare approaches", "benchmark against my features", "prove GAFIME works", or "GAFIME vs hand-crafted".

Generate a complete scikit-learn pipeline with GAFIME feature interaction mining. Use when the user wants to create an ML pipeline, integrate GAFIME with sklearn, build a classification or regression workflow, or says things like "build a pipeline", "create a model with GAFIME", "sklearn integration", "churn prediction pipeline", "set up classification", or "generate training script".

Profile and analyze a dataset before running GAFIME feature interaction mining. Use when the user wants to know if their data is compatible with GAFIME, estimate VRAM requirements, check for problematic columns, determine optimal batch size, or asks things like "will my data fit on GPU", "profile my dataset", "analyze my CSV/Parquet", "how many features do I have", or "is my data ready for GAFIME".

Configure GAFIME Engine v0.4.x time-series candidate search for temporal/sequential data. Use when the user has temporal data, transaction logs, time-stamped records, or asks things like "set up time series features", "time series feature engineering", "velocity features", "rolling windows", or "regime features".

Validate whether GAFIME-discovered feature interactions are genuinely predictive or just noise. Use when the user wants to verify feature quality, test for overfitting, check out-of-sample performance of discovered interactions, or says things like "are these features real", "validate features", "check for overfitting", "out-of-sample test", "are these interactions genuine", or "test feature significance".

Interpret and explain GAFIME DiagnosticReport results in plain English. Use when the user has run GAFIME and wants to understand the output, asks about Pearson correlation values, p-values, stability metrics, signal detection, feature interaction results, or says things like "what does this report mean", "explain the results", "is this feature interaction significant", "what does signal detected mean", or "interpret my GAFIME output".