API Reference

The Python library is the source of truth. The REST layer is a thin 1:1 wrapper: zero logic, exceptions mapped to HTTP codes (LookupError → 404, ValueError → 400, FileNotFoundError → 404).

endpoint	library function
`GET /tickers?sector=&industry=`	`list_tickers(session, sector?, industry?)`
`GET /tickers/{code}/daily?start=&end=`	`get_daily_candles(session, code, start?, end?)`
`GET /tickers/{code}/metrics?metrics=&asof_lag=`	`get_metrics(session, code, metrics?, start?, end?, asof_lag?)`
`GET /panel/{metric}?start=&end=`	`get_panel(metric, start?, end?)`
`GET /metrics`	`list_metrics()`

Library

from db_interface import SessionLocal
from market_api import get_daily_candles, get_metrics, get_panel, list_metrics, list_tickers

session = SessionLocal()

list_tickers(session, sector="IT")          # reference, filterable
get_daily_candles(session, "INFY", start=date(2024, 1, 1))
get_metrics(session, "INFY", metrics=["mom1m", "mvel1"], asof_lag=3)
get_panel("mom1m")                          # wide date × ticker cross-section
list_metrics()                              # metric | category | group | description

REST

python -m uvicorn market_api.app:app --port 8000

curl "localhost:8000/tickers?sector=IT"
curl "localhost:8000/tickers/INFY/daily?start=2024-01-01"
curl "localhost:8000/tickers/INFY/metrics?metrics=mom1m,mvel1&asof_lag=3"
curl "localhost:8000/panel/mom1m"
curl "localhost:8000/metrics"

Responses are JSON with DataFrames in records orientation; NaN values are serialized as null.

Errors

status	meaning
404	unknown ticker, or metric not found in any panel
400	unknown metric name in `metrics=`

CLI

python -m market_api.cli export [--nse-code INFY]   # stage per-ticker CSVs
python -m market_api.cli build-panel [--all]        # rebuild panel CSVs

Home

Concepts

API Reference

Library

REST

Errors

CLI

​Library

​REST

​Errors

​CLI

Library

REST

Errors

CLI