clj-yfinance.dataset — com.github.clojure-finance/clj-yfinance 0.1.9

Optional dataset conversion functions for integration with the Clojure data science ecosystem.

REQUIRES: techascent/tech.ml.dataset on the classpath.

The results are plain tech.v3.dataset datasets, so they work directly with tablecloth, Noj, datajure and anything else built on tech.ml.dataset.

EXAMPLES:

(require '[clj-yfinance.dataset :as yfd]) (yfd/historical->dataset "AAPL" :period "1mo") (yfd/multi-ticker->dataset ["AAPL" "GOOGL" "MSFT"] :period "1y")

;; With tablecloth (require '[tablecloth.api :as tc]) (-> (yfd/historical->dataset "AAPL" :period "1mo") (tc/add-column :returns (fn [ds] ...)) (tc/select-columns [:timestamp :close :returns]))

dividends-splits->dataset clj

(dividends-splits->dataset data-or-ticker)

(dividends-splits->dataset ticker & opts)

Convert dividends and splits data to separate datasets.

Returns a map with :dividends and :splits keys, each containing a dataset; a key is absent when there are no events of that kind. Returns nil if both are empty. Timestamps come from each event's :date; rows are in date order.

Dividend columns: :timestamp (int64), :amount (float64) Split columns: :timestamp (int64), :numerator (float64), :denominator (float64)

Example: (def data (yf/fetch-dividends-splits "AAPL" :period "10y")) (dividends-splits->dataset data) ;; => {:dividends #tech.v3.dataset<...>, :splits #tech.v3.dataset<...>}

(dividends-splits->dataset "AAPL" :period "5y")

historical->dataset clj

(historical->dataset data-or-ticker)

(historical->dataset ticker & opts)

Convert historical OHLCV data to tech.v3.dataset with proper column types.

Accepts either:

• (ticker & opts) - fetches data then converts
• (data-map) - converts already-fetched data

Returns tech.v3.dataset with typed columns:

• :timestamp (int64) - Unix epoch seconds
• :open, :high, :low, :close, :adj-close (float64)
• :volume (int64)

Returns nil if data is empty or fetch fails.

Examples: (historical->dataset "AAPL" :period "1mo") (historical->dataset "TSLA" :period "1y" :interval "1d")

;; Convert already-fetched data (def data (yf/fetch-historical "AAPL" :period "1mo")) (historical->dataset data)

info->dataset clj

(info->dataset data-or-ticker)

(info->dataset ticker & _opts)

Convert ticker info to single-row dataset.

Returns a single-row dataset containing ticker metadata. Returns nil if fetch fails.

Useful for combining multiple ticker info into a comparison table:

Example: (info->dataset "AAPL")

;; Compare multiple tickers (require '[tech.v3.dataset :as ds]) (apply ds/concat (map info->dataset ["AAPL" "GOOGL" "MSFT"]))

multi-ticker->dataset clj

(multi-ticker->dataset tickers & opts)

Fetch historical data for multiple tickers and combine into single dataset.

Returns a dataset with all tickers combined, including a :ticker column to identify each row's source ticker.

Returns nil if all fetches fail or all data is empty.

Columns: :ticker (string), :timestamp (int64), :open/:high/:low/:close/:adj-close (float64), :volume (int64)

Example: (multi-ticker->dataset ["AAPL" "GOOGL" "MSFT"] :period "1y") (multi-ticker->dataset ["AAPL" "TSLA"] :period "1mo" :interval "1h")

;; Use with tablecloth for grouped operations (require '[tablecloth.api :as tc]) (-> (multi-ticker->dataset ["AAPL" "GOOGL" "MSFT"] :period "1y") (tc/group-by [:ticker]) (tc/add-column :returns (fn [ds] ...)))

prices->dataset clj

(prices->dataset prices-map)

Convert price map to dataset with ticker and price columns.

Takes a map of {ticker price} (as returned by fetch-prices) and converts to a dataset with :ticker (string) and :price (float64) columns.

Returns nil if prices map is empty.

Example: (def prices (yf/fetch-prices ["AAPL" "GOOGL" "MSFT"])) (prices->dataset prices) ;; => tech.v3.dataset with columns [:ticker :price]