Market data by id

Through the REST API you don't need to bring the data for a market. Search it, take its id, and pass the id to the analyses. This works in every plan.

The MCP server takes no market data. It analyses the data you give it: values from another connector (a market-data connector, for example), a file, your own numbers, or your stored datasets. The market search and the analysis by id below are REST only.

The flow

1. Search. GET /api/data/SearchSymbols?search=Apple. Every hit carries a symbolId.
2. Analyse by id. Pass the id where an analysis takes a dataset: - datasetid=<symbolId> on the cycle explorer, the cycle scanner, the cyclic RSI, the consensus calculation, the cycle composite and the power scanner; - GET /api/cycles/MarketCycles/...: the dominant cycle of a market in one call; - GET /api/CycleConsensus/score/...: the consensus of a market in one call; - the DTW clustering (/api/dtw): 2 to 20 ids at once, to see which markets move alike.
3. Read the cycles. The answer carries the cycles, their phases, the projections, the cycle highlighter and the CRSI, as it does for your own data. See Concepts.

Use the ids exactly as the search returns them.

Window and freshness

The API brings a market's data up to date before it analyses it, at most once every six hours per symbol. A market is analysed over its last 1,250 days unless maxbars or from say otherwise, and every answer names what was analysed (analysedWindow: the id, the first and the last bar's date, the number of bars).

What the answers contain, and what they leave out

The analyses return what the model measures: cycles, phases, projections, oscillators and scores. They do not return the prices of a market. Without the feature MarketDataAccess, answers about market data leave out the price fields: currentPrice, price, smoothedPrice, lastInputValue, preprocessedValues and indexedPerformance. The cycle highlighter, the dominant cycle wave and the CRSI stay in.

For your own data nothing is left out. You sent the values, so you get them back.

What needs MarketDataAccess

The raw bars themselves are not part of any plan:

• every route under /api/data/ except the symbol search;
• smoothing, detrending and the turning-point test with a market id: their answer is the transformed series itself. With your own data they are open in every plan.

The feature is granted on request, for a project that needs market data by key. Without it these calls are answered 403 with a message that says so and names the way to analyse by id instead.

Your own datasets and streams

datasetid=<name> also names the datasets you stored with your key (PUT /api/datasets/{name}, or store_dataset through the MCP), the datasets you uploaded in the app and your own live streams. Datasets of other users are not open: a call on another user's feed is refused with 403.

Limits

A market analysis counts once on your account, like every other call. The clustering takes 2 to 20 datasets, up to 5 cluster counts from 2 to 10, at most 10 repetitions and 1,000 bars back. See Plans and limits.

Answers of Cycles IQ are measurements and projections of cycles, not investment advice. Questions: support.