How kkritika123/solar-flare-forecast works
Solar Flare Forecast retrieves historical CCMC forecasts and LMSAL observations, normalizes forecasts, then evaluates forecast windows using TSS/HSS. It also offers direct HAPI and processed-archive queries and source comparison. The principal archive path runs through the public API and processing pipeline; evaluation is separately orchestrated by the API. Edges reflect shown calls, SOURCE INDEX references, or README statements; unsampled implementation details are omitted.
Main entry points: scoring.py Evaluation orchestration · api.py Public API · pipeline.py Download pipeline · query_forecasts.py Query command
Components 23
Forecast ingestion
- Public API
flare_scoreboard/api.py - Workflow config
flare_scoreboard/config.py - Model discovery
flare_scoreboard/crawl.py - Archive HTTP
flare_scoreboard/http_client.py - Download pipeline
flare_scoreboard/pipeline.py - Forecast parsers
flare_scoreboard/parsers.py - Parse normalization
flare_scoreboard/parse_core.py - Yearly CSV output
flare_scoreboard/csv_output.py - Processed forecasts
Forecast queries
- Query command
query_forecasts.py - Query and compare
flare_scoreboard/api.py - HAPI client
flare_scoreboard/hapi.py - CCMC HAPI
- Query CSV results
data
Observed events
- LMSAL retrieval
flare_scoreboard/lmsal.py - LMSAL archive
- Observed event data
data
Evaluation
- Evaluation orchestration
flare_scoreboard/scoring.py - Window matching and scores
flare_scoreboard/evaluation.py - Evaluation CSV results
- Legacy evaluation script
model.py - Evaluation helpers
flare_eval_utils.py
Other components
- Researcher
Connections 28
- Researcher → Public API: requests download/evaluation
README.md - Public API → Workflow config: loads settings
flare_scoreboard/api.py - Public API → Model discovery: discovers models
flare_scoreboard/api.py - Public API → Download pipeline: processes models
flare_scoreboard/api.py - Download pipeline → Model discovery: crawls files
flare_scoreboard/pipeline.py - Download pipeline → Archive HTTP: downloads files
flare_scoreboard/pipeline.py - Download pipeline → Forecast parsers: parses files
flare_scoreboard/pipeline.py - Forecast parsers → Parse normalization: uses helpers
flare_scoreboard/parsers.py - Download pipeline → Yearly CSV output: writes yearly rows
flare_scoreboard/pipeline.py - Yearly CSV output → Processed forecasts: writes forecasts
README.md - Researcher → Query command: runs query
README.md - Query command → Query and compare: queries archive
query_forecasts.py - Query command → HAPI client: queries HAPI
query_forecasts.py - Query and compare → HAPI client: requests predictions
flare_scoreboard/api.py - HAPI client → CCMC HAPI: requests data
flare_scoreboard/hapi.py - Query command → Processed forecasts: reads archive CSVs
README.md - Query command → Query CSV results: saves query CSVs
README.md - Researcher → LMSAL retrieval: requests event download
README.md - LMSAL retrieval → LMSAL archive: retrieves events
README.md - LMSAL retrieval → Observed event data: writes event CSV
README.md - Public API → Evaluation orchestration: runs evaluation
flare_scoreboard/api.py - Evaluation orchestration → Window matching and scores: calculates scores
flare_scoreboard/scoring.py - Evaluation orchestration → Processed forecasts: reads forecasts
README.md - Evaluation orchestration → Observed event data: reads observations
README.md - Evaluation orchestration → LMSAL retrieval: downloads if missing
README.md - Evaluation orchestration → Evaluation CSV results: writes scores
README.md - Researcher → Legacy evaluation script: runs legacy workflow
README.md - Legacy evaluation script → Evaluation helpers: uses helpers
model.py