Add ziplime PIT dataset: docs, manifest, recipe, ingest, workflow, bundle

.gitattributes

@@ -1,60 +1,2 @@
-*.7z filter=lfs diff=lfs merge=lfs -text
-*.arrow filter=lfs diff=lfs merge=lfs -text
-*.avro filter=lfs diff=lfs merge=lfs -text
-*.bin filter=lfs diff=lfs merge=lfs -text
-*.bz2 filter=lfs diff=lfs merge=lfs -text
-*.ckpt filter=lfs diff=lfs merge=lfs -text
-*.ftz filter=lfs diff=lfs merge=lfs -text
-*.gz filter=lfs diff=lfs merge=lfs -text
-*.h5 filter=lfs diff=lfs merge=lfs -text
-*.joblib filter=lfs diff=lfs merge=lfs -text
-*.lfs.* filter=lfs diff=lfs merge=lfs -text
-*.lz4 filter=lfs diff=lfs merge=lfs -text
-*.mds filter=lfs diff=lfs merge=lfs -text
-*.mlmodel filter=lfs diff=lfs merge=lfs -text
-*.model filter=lfs diff=lfs merge=lfs -text
-*.msgpack filter=lfs diff=lfs merge=lfs -text
-*.npy filter=lfs diff=lfs merge=lfs -text
-*.npz filter=lfs diff=lfs merge=lfs -text
-*.onnx filter=lfs diff=lfs merge=lfs -text
-*.ot filter=lfs diff=lfs merge=lfs -text
 *.parquet filter=lfs diff=lfs merge=lfs -text
-*.pb filter=lfs diff=lfs merge=lfs -text
-*.pickle filter=lfs diff=lfs merge=lfs -text
-*.pkl filter=lfs diff=lfs merge=lfs -text
-*.pt filter=lfs diff=lfs merge=lfs -text
-*.pth filter=lfs diff=lfs merge=lfs -text
-*.rar filter=lfs diff=lfs merge=lfs -text
-*.safetensors filter=lfs diff=lfs merge=lfs -text
-saved_model/**/* filter=lfs diff=lfs merge=lfs -text
-*.tar.* filter=lfs diff=lfs merge=lfs -text
-*.tar filter=lfs diff=lfs merge=lfs -text
-*.tflite filter=lfs diff=lfs merge=lfs -text
-*.tgz filter=lfs diff=lfs merge=lfs -text
-*.wasm filter=lfs diff=lfs merge=lfs -text
-*.xz filter=lfs diff=lfs merge=lfs -text
-*.zip filter=lfs diff=lfs merge=lfs -text
-*.zst filter=lfs diff=lfs merge=lfs -text
-*tfevents* filter=lfs diff=lfs merge=lfs -text
-# Audio files - uncompressed
-*.pcm filter=lfs diff=lfs merge=lfs -text
-*.sam filter=lfs diff=lfs merge=lfs -text
-*.raw filter=lfs diff=lfs merge=lfs -text
-# Audio files - compressed
-*.aac filter=lfs diff=lfs merge=lfs -text
-*.flac filter=lfs diff=lfs merge=lfs -text
-*.mp3 filter=lfs diff=lfs merge=lfs -text
-*.ogg filter=lfs diff=lfs merge=lfs -text
-*.wav filter=lfs diff=lfs merge=lfs -text
-# Image files - uncompressed
-*.bmp filter=lfs diff=lfs merge=lfs -text
-*.gif filter=lfs diff=lfs merge=lfs -text
-*.png filter=lfs diff=lfs merge=lfs -text
-*.tiff filter=lfs diff=lfs merge=lfs -text
-# Image files - compressed
-*.jpg filter=lfs diff=lfs merge=lfs -text
-*.jpeg filter=lfs diff=lfs merge=lfs -text
-*.webp filter=lfs diff=lfs merge=lfs -text
-# Video files - compressed
-*.mp4 filter=lfs diff=lfs merge=lfs -text
-*.webm filter=lfs diff=lfs merge=lfs -text
+data.delta/** filter=lfs diff=lfs merge=lfs -text

.github/workflows/update.yml

@@ -0,0 +1,27 @@
+name: update-finance-yahoo-data
+
+# every US trading day, shortly after the 16:00 ET close
+on:
+  schedule:
+    - cron: "0 22 * * 1-5"
+  workflow_dispatch:
+
+jobs:
+  ingest:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          lfs: true
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+      - run: pip install polars deltalake httpx huggingface_hub
+      - name: Fetch + append (PIT, append-only)
+        run: python ingest.py --since "$(date -u -d '14 days ago' +%F)"
+      - name: Push updated bundle
+        env:
+          HF_TOKEN: ${{ secrets.HF_TOKEN }}
+        run: |
+          huggingface-cli upload --repo-type dataset \
+            ZipLime/finance-yahoo-data data/ data/ --token "$HF_TOKEN"

README.md

@@ -0,0 +1,158 @@
+---
+license: other
+language:
+  - en
+pretty_name: Yahoo Equity Daily Bars (PIT)
+tags:
+  - point-in-time
+  - pit
+  - ziplime
+  - backtesting
+  - alternative-data
+  - finance
+  - us_equities
+task_categories:
+  - time-series-forecasting
+size_categories:
+  - n<1K
+configs:
+  - config_name: default
+    data_files:
+      - split: train
+        path: data/data_bundle/**/*.parquet
+---
+
+# 📈 Yahoo Equity Daily Bars (PIT)
+
+Daily OHLCV bars for US equities, delivered as a point-in-time ziplime bundle.
+
+Part of the **ziplime Point-in-Time (PIT) data layer** — append-only datasets with an
+explicit split between when a fact *happened* (`event_date`) and when it became *known*
+(`knowledge_date`). A simulation at time **T** can only ever observe rows with
+`knowledge_date <= T`, so restatements, publication lag and hindsight can't leak into a
+backtest. The identical code path runs live with **T = now**.
+
+- **Data class:** Market data — daily equity OHLCV
+- **Entity domain:** `us_equities` — US-listed equity, keyed by ticker (resolves via the `entity_map` dataset).
+- **Origin:** Yahoo Finance (`yfinance`), auto-adjusted daily bars
+- **License:** Yahoo Finance terms — research use only
+- **Update cadence:** every US trading day, shortly after the 16:00 ET close (`0 22 * * 1-5`)
+- **Format:** ziplime Delta Lake bundle (`data_type: PIT_DATA`)
+
+## Why point-in-time?
+
+Backtests on non-price data are systematically optimistic when the data layer has no notion
+of *when a fact became known*. Three failure modes this dataset is built to avoid:
+
+1. **Restatements** — a value reported one quarter and revised the next. Storing only the
+   final value lets a backtest "know" the revision months early.
+2. **Publication lag** — fundamentals keyed by fiscal-period-end, joined to prices at
+   period end rather than the (weeks-later) filing date.
+3. **Hindsight in derived signals** — a recent model scoring old text has already seen how
+   the story ended.
+
+All three are the same bug, and it is fixed in the data layer, not in strategy code.
+
+## Schema
+
+### System columns (every PIT dataset)
+
+| Column | Type | Semantics |
+|---|---|---|
+| `entity_id` | Utf8 | Stable entity identifier (resolved via the `entity_map` PIT dataset) |
+| `event_date` | Timestamp(UTC, µs) | The moment the fact refers to |
+| `knowledge_date` | Timestamp(UTC, µs) | The moment it became publicly known — **the only column the as-of filter uses** |
+| `knowledge_estimated` | Boolean | `true` if `knowledge_date` was reconstructed by a lag model rather than taken from the source |
+| `ingested_at` | Timestamp(UTC, µs) | When our pipeline wrote the row (audit only; never used in as-of) |
+
+### Value columns (this dataset)
+
+| Column | Type | Description |
+|---|---|---|
+| `open` | Float64 | Session open (auto-adjusted) |
+| `high` | Float64 | Session high |
+| `low` | Float64 | Session low |
+| `close` | Float64 | Session close (auto-adjusted) |
+| `volume` | Float64 | Session volume |
+| `price` | Float64 | Convenience alias of close |
+
+The logical key of a fact is `(entity_id, event_date)`. A **revision** is a new row with the
+same key and a later `knowledge_date`. Written rows are immutable; history is never rewritten.
+
+## As-of access
+
+Inside a ziplime strategy there is **no `T` parameter** — the knowledge moment always equals
+the simulation clock (live: wall clock):
+
+```python
+async def initialize(context):
+    context.ds = await context.pit("finance-yahoo-data")
+
+async def handle_data(context, data):
+    # only rows with knowledge_date <= current simulation time are visible
+    latest = await context.ds.latest(
+        assets=[context.asset], fields=['open', 'high']
+    )
+    history = await context.ds.as_of(
+        assets=[context.asset], fields=['open'],
+        event_range=("2022-01-01", None),
+    )
+```
+
+### Reading it outside ziplime (plain Polars + delta-rs)
+
+```python
+import polars as pl
+
+T = "2025-06-01T00:00:00Z"          # "what was known at T"
+lf = pl.scan_delta("hf://datasets/ZipLime/finance-yahoo-data/data/data_bundle/yahoo_finance_daily_data/1784755946/data.delta")
+as_of = (
+    lf.filter(pl.col("knowledge_date") <= T)
+      .sort("knowledge_date")
+      .group_by(["entity_id", "event_date"], maintain_order=True)
+      .last()
+)
+print(as_of.collect())
+```
+
+Delta time-travel (`AS OF <version>`) pins the table for reproducibility; the
+`knowledge_date <= T` filter is what enforces point-in-time. They compose: a backtest records
+`(dataset, delta_version)` and replays read the table at that version *and* apply the filter.
+
+## Updates
+
+`recipe.py` implements the collection contract `fetch(since: datetime) -> pl.DataFrame` in the
+PIT schema above; `ingest.py` dedups and **appends** to the Delta bundle (never rewrites).
+The scheduled job in `.github/workflows/update.yml` runs it every US trading day, shortly after the 16:00 ET close.
+
+```python
+# recipe.py (contract)
+async def fetch(since: datetime) -> "pl.DataFrame": ...
+```
+
+## Knowledge-date convention
+
+For quotes `knowledge_date ≈ event_date`: a daily bar is only complete once its session closes, so it is stamped visible from the close of that session. OHLCV is the one class the design doc keeps out of full PIT treatment — this dataset is the bootstrap bundle that seeds the format.
+
+## What's in this repo
+
+```
+README.md                     # this card
+manifest.json                 # PIT dataset manifest (schema, source, schedule)
+recipe.py                     # fetch(since) -> PIT rows
+ingest.py                     # dedup + append-only Delta writer
+.github/workflows/update.yml  # scheduled ingestion
+data/                         # ziplime Delta bundle + registry manifest
+  bundle_registry/yahoo_finance_daily_data_1784755946.json
+  data_bundle/yahoo_finance_daily_data/1784755946/data.delta/
+```
+
+The `data/` bundle is a ready-to-load ziplime Delta Lake market-data bundle (five US equity
+tickers, daily bars) that seeds the pipeline and lets you exercise the loader end-to-end
+today. Point `pl.scan_delta` (above) at it, or register it with ziplime's
+`FileSystemBundleRegistry`.
+
+---
+
+<sub>Generated for the ziplime PIT data-layer prototype. Manifest and schema follow the
+ziplime PIT spec; `source.*` fields declare origin and license per the dataset manifest.</sub>

data/bundle_registry/yahoo_finance_daily_data_1784755946.json

@@ -0,0 +1,13 @@
+{
+  "name": "yahoo_finance_daily_data",
+  "version": "1784755946",
+  "bundle_storage_class": "ziplime.data.services.file_system_delta_lake_bundle_storage.FileSystemDeltaLakeBundleStorage",
+  "bundle_storage_data": {
+    "base_data_path": "/Users/vyacheslav/.ziplime/data"
+  },
+  "trading_calendar_name": "XNYS",
+  "frequency_seconds": 86400.0,
+  "frequency_text": null,
+  "timestamp": "2026-07-22T17:32:32Z",
+  "data_type": "MARKET_DATA"
+}

data/data_bundle/yahoo_finance_daily_data/1784755946/data.delta/_delta_log/00000000000000000000.json

@@ -0,0 +1,4 @@
+{"commitInfo":{"timestamp":1784755953145,"operation":"WRITE","operationParameters":{"mode":"ErrorIfExists"},"engineInfo":"delta-rs:py-1.6.2","operationMetrics":{"execution_time_ms":11,"num_added_files":1,"num_added_rows":825,"num_partitions":0,"num_removed_files":0},"clientVersion":"delta-rs.py-1.6.2"}}
+{"protocol":{"minReaderVersion":1,"minWriterVersion":2}}
+{"metaData":{"id":"3ddbf1f1-f8a3-4e40-a0be-751151e72bc4","name":null,"description":null,"format":{"provider":"parquet","options":{}},"schemaString":"{\"type\":\"struct\",\"fields\":[{\"name\":\"date\",\"type\":\"timestamp\",\"nullable\":true,\"metadata\":{}},{\"name\":\"open\",\"type\":\"double\",\"nullable\":true,\"metadata\":{}},{\"name\":\"high\",\"type\":\"double\",\"nullable\":true,\"metadata\":{}},{\"name\":\"low\",\"type\":\"double\",\"nullable\":true,\"metadata\":{}},{\"name\":\"close\",\"type\":\"double\",\"nullable\":true,\"metadata\":{}},{\"name\":\"volume\",\"type\":\"double\",\"nullable\":true,\"metadata\":{}},{\"name\":\"symbol\",\"type\":\"string\",\"nullable\":true,\"metadata\":{}},{\"name\":\"mic\",\"type\":\"string\",\"nullable\":true,\"metadata\":{}},{\"name\":\"price\",\"type\":\"double\",\"nullable\":true,\"metadata\":{}},{\"name\":\"sid\",\"type\":\"long\",\"nullable\":true,\"metadata\":{}},{\"name\":\"backfilled\",\"type\":\"boolean\",\"nullable\":true,\"metadata\":{}}]}","partitionColumns":[],"createdTime":1784755953134,"configuration":{}}}
+{"add":{"path":"part-00000-d6922c4e-1b91-4a1a-b856-d849d69a8f52-c000.snappy.parquet","partitionValues":{},"size":44249,"modificationTime":1784755953145,"dataChange":true,"stats":"{\"numRecords\":825,\"minValues\":{\"low\":82.11000061035156,\"close\":82.83999633789062,\"mic\":\"XNMS\",\"sid\":4470,\"volume\":6743500.0,\"symbol\":\"AAPL\",\"backfilled\":false,\"high\":84.48899841308594,\"open\":82.78500366210938,\"date\":\"2025-01-02T05:00:00Z\",\"price\":82.83999633789062},\"maxValues\":{\"close\":787.419189453125,\"date\":\"2025-08-29T04:00:00Z\",\"price\":787.419189453125,\"symbol\":\"NFLX\",\"sid\":10977,\"backfilled\":false,\"mic\":\"XNMS\",\"low\":778.269228256533,\"volume\":184395900.0,\"high\":793.6488143059948,\"open\":788.5653976304532},\"nullCount\":{\"open\":0,\"sid\":0,\"volume\":0,\"symbol\":0,\"backfilled\":0,\"date\":0,\"price\":0,\"close\":0,\"mic\":0,\"low\":0,\"high\":0}}","tags":null,"baseRowId":null,"defaultRowCommitVersion":null,"clusteringProvider":null}}

data/data_bundle/yahoo_finance_daily_data/1784755946/data.delta/part-00000-d6922c4e-1b91-4a1a-b856-d849d69a8f52-c000.snappy.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45bd396c116f8b9ff35650c9fc88b4dd30465e2421b1ac8f91456310a56513e9
+size 44249

ingest.py

@@ -0,0 +1,57 @@
+"""
+Shared append-only ingest harness for a ziplime PIT dataset.
+
+Reads `recipe.fetch(since)`, stamps `ingested_at`, drops byte-identical duplicates of
+existing (entity_id, event_date, knowledge_date) rows, and **appends** to the Delta table.
+It never issues UPDATE/DELETE/MERGE — append-only is the PIT invariant the linter enforces.
+"""
+from __future__ import annotations
+import argparse, hashlib
+from datetime import datetime, timezone
+from pathlib import Path
+
+import polars as pl
+from deltalake import DeltaTable, write_deltalake
+
+import recipe
+
+TABLE_URI = str(Path(__file__).parent / "data" / "data_bundle" /
+                "yahoo_finance_daily_data" / "1784755946" / "data.delta")
+
+
+def _row_hash(df: pl.DataFrame) -> pl.Series:
+    cols = [c for c in df.columns if c != "ingested_at"]
+    return df.select(cols).hash_rows().cast(pl.Utf8)
+
+
+async def main(since: datetime) -> None:
+    fresh = await recipe.fetch(since)
+    if fresh.is_empty():
+        print("recipe returned no rows; nothing to append")
+        return
+
+    fresh = fresh.with_columns(
+        pl.lit(datetime.now(timezone.utc)).alias("ingested_at"),
+    )
+
+    try:
+        existing = pl.from_arrow(DeltaTable(TABLE_URI).to_pyarrow_table())
+        seen = set(_row_hash(existing).to_list())
+        fresh = fresh.filter(~_row_hash(fresh).is_in(seen))
+    except Exception:
+        pass  # first ingest / fresh table
+
+    if fresh.is_empty():
+        print("no new or revised rows after dedup")
+        return
+
+    write_deltalake(TABLE_URI, fresh.to_arrow(), mode="append")
+    print(f"appended {len(fresh)} rows to {TABLE_URI}")
+
+
+if __name__ == "__main__":
+    ap = argparse.ArgumentParser()
+    ap.add_argument("--since", default="2025-01-01")
+    args = ap.parse_args()
+    import asyncio
+    asyncio.run(main(datetime.fromisoformat(args.since).replace(tzinfo=timezone.utc)))

manifest.json

@@ -0,0 +1,31 @@
+{
+  "name": "finance_yahoo_data",
+  "version": "1784755946",
+  "data_type": "PIT_DATA",
+  "bundle_storage_class": "ziplime.data.services.file_system_delta_lake_bundle_storage.FileSystemDeltaLakeBundleStorage",
+  "bundle_storage_data": {
+    "table_uri": "data/data_bundle/yahoo_finance_daily_data/1784755946/data.delta"
+  },
+  "entity_domain": "us_equities",
+  "schema": {
+    "open": "Float64",
+    "high": "Float64",
+    "low": "Float64",
+    "close": "Float64",
+    "volume": "Float64",
+    "price": "Float64"
+  },
+  "pit": {
+    "knowledge_lag_model": null,
+    "model_knowledge_cutoff": null
+  },
+  "source": {
+    "recipe": "recipe.py",
+    "recipe_hash": "sha256:<filled-by-ci>",
+    "origin": "Yahoo Finance (`yfinance`), auto-adjusted daily bars",
+    "license": "Yahoo Finance terms — research use only",
+    "schedule": "0 22 * * 1-5"
+  },
+  "trading_calendar_name": null,
+  "frequency_seconds": null
+}

recipe.py

@@ -0,0 +1,49 @@
+"""
+Yahoo Equity Daily Bars (PIT) — collection recipe.
+
+Contract (ziplime PIT recipe, design doc §9):
+
+    async def fetch(since: datetime) -> pl.DataFrame
+
+Returns rows in the PIT schema for `finance-yahoo-data`:
+  system:  entity_id, event_date, knowledge_date, knowledge_estimated
+  values:  open, high, low, close, volume, price
+
+The shared harness (ingest.py) stamps `ingested_at`, dedups against existing rows and
+appends to the Delta bundle. This recipe only fetches and shapes.
+"""
+from __future__ import annotations
+from datetime import datetime, timezone
+
+
+async def fetch(since: datetime):
+    import polars as pl  # noqa: F401
+    import yfinance as yf
+    import polars as pl
+
+    universe = ["META", "AAPL", "AMZN", "NFLX", "GOOGL"]
+    raw = yf.download(
+        tickers=universe, interval="1d", start=since, threads=1,
+        group_by="Ticker", auto_adjust=True, multi_level_index=False, progress=False,
+    )
+    frames = []
+    for sym in universe:
+        d = pl.from_pandas(raw[sym], include_index=True).rename({
+            "Date": "event_date", "Open": "open", "High": "high",
+            "Low": "low", "Close": "close", "Volume": "volume",
+        })
+        d = d.with_columns(
+            pl.lit(sym).alias("entity_id"),
+            # a daily bar is known at its own session close
+            pl.col("event_date").alias("knowledge_date"),
+            pl.lit(False).alias("knowledge_estimated"),
+            pl.col("close").alias("price"),
+        )
+        frames.append(d)
+    return pl.concat(frames)
+
+
+if __name__ == "__main__":
+    import asyncio, polars as pl
+    df = asyncio.run(fetch(datetime(2025, 1, 1, tzinfo=timezone.utc)))
+    print(df.head())