simudyne-pulse 0.7.0.dev2__tar.gz → 0.7.0.dev4__tar.gz

{simudyne_pulse-0.7.0.dev2 → simudyne_pulse-0.7.0.dev4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: simudyne-pulse
-Version: 0.7.0.dev2
+Version: 0.7.0.dev4
 Summary: Python SDK for the Simudyne Pulse synthetic market data API
 Author-email: Simudyne <support@simudyne.com>
 License-Expression: MIT

{simudyne_pulse-0.7.0.dev2 → simudyne_pulse-0.7.0.dev4}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "simudyne-pulse"
-version = "0.7.0-dev.2"
+version = "0.7.0-dev.4"
 description = "Python SDK for the Simudyne Pulse synthetic market data API"
 readme = "README.md"
 license = "MIT"

{simudyne_pulse-0.7.0.dev2 → simudyne_pulse-0.7.0.dev4}/src/simudyne/resources/data.py

@@ -11,6 +11,7 @@ class DataResource:
         exchange: str | None = None,
         provider: str | None = None,
         date: str | None = None,
+        q: str | None = None,
         limit: int | None = None,
         offset: int | None = None,
     ):
@@ -26,6 +27,7 @@ class DataResource:
             date: Calibration date "YYYY-MM-DD". Keeps only instruments
                   calibrated on that date, and narrows each instrument's
                   available_dates to it.
+            q: Case-insensitive substring match over ticker and company name.
             limit: Max instruments to return.
             offset: Instruments to skip (for paging alongside limit).
 
@@ -38,6 +40,7 @@ class DataResource:
                 "exchange": exchange,
                 "provider": provider,
                 "date": date,
+                "q": q,
                 "limit": limit,
                 "offset": offset,
             }.items() if v is not None

{simudyne_pulse-0.7.0.dev2 → simudyne_pulse-0.7.0.dev4}/src/simudyne/resources/simulation.py

@@ -299,13 +299,16 @@ class SimulationResource:
             payload["batch_size"] = batch_size
         return self._pro_request("POST", CALIBRATE_PATH, json=payload)
 
-    def get_jobs(self):
+    def get_jobs(self, limit: int = 100):
         """
-        Get all simulation jobs submitted by the authenticated user.
+        Get simulation jobs submitted by the authenticated user, newest first.
         
         Returns a list of jobs with their associated simulation IDs. Use this
         to find job IDs for past runs or to see what simulations are pending.
         
+        Args:
+            limit: Max jobs to return (default 100, max 500)
+        
         Returns:
             dict: Contains:
                 - jobs (list): List of job objects, each with:
@@ -313,6 +316,7 @@ class SimulationResource:
                     - sim_ids (list): List of simulation IDs in this job
                     - created_at (str): Timestamp when job was submitted
                 - total (int): Total number of jobs
+                - returned (int): Jobs in this page
                 
         Example:
             >>> result = client.simulation.get_jobs()
@@ -322,7 +326,7 @@ class SimulationResource:
             ...     print(f"Job {job['job_id']}: {len(job['sim_ids'])} simulations")
             ...     print(f"  Created: {job['created_at']}")
         """
-        return self._pro_request("GET", JOBS_PATH)
+        return self._pro_request("GET", JOBS_PATH, params={"limit": limit})
 
     def get_job_status(self, job_id: str):
         """

{simudyne_pulse-0.7.0.dev2 → simudyne_pulse-0.7.0.dev4}/src/simudyne/resources/validation.py

@@ -53,11 +53,56 @@ _TRI_STATE_FLAGS = (
     "plot_data",
 )
 
+#: Areas of checking selectable per job, as of pulse-check 1.10.0. Left unset
+#: they take the server's default, so a job that names none behaves as before.
+_AREA_FLAGS = ("statistical", "stylised_facts", "impact", "volume_correlation",
+               "fid", "mind")
+
+#: Everything else the 1.10.0 schema accepts. ``lob`` marks the frames as L2
+#: snapshots, which switches off anything needing the message stream;
+#: ``sample_period`` and ``match_generated_sample`` set the grid the book is
+#: resampled onto; ``plots`` is False, True, or a list of plot ids;
+#: ``historical_output`` is the demo-only gate that ``plot_data`` used to be.
+_EXTRA_FIELDS = ("lob", "sample_period", "match_generated_sample", "plots",
+                 "historical_output")
+
 #: SDK name -> API config field. The API kept ``run_fid`` for compatibility;
 #: the SDK spells out what it actually gates.
 _INCEPTION_WIRE_FIELD = "run_fid"
 
 
+def _frame_to_parquet(entry, index: int):
+    """Normalise one simulated run to ``(filename, parquet bytes)``.
+
+    Accepts a path, a ``(filename, bytes)`` pair, or a polars / pandas
+    DataFrame, which is written to parquet in memory. The frame must be pulse
+    format -- the same shape the engine writes to ``sim_data.parquet`` -- which
+    the server validates; sending something else fails there, not here.
+    """
+    import io
+
+    if isinstance(entry, tuple):
+        return entry
+
+    # Duck-typed rather than imported: neither polars nor pandas is a hard
+    # dependency of the SDK, and importing one to test for the other would
+    # make it one.
+    writer = getattr(entry, "write_parquet", None)      # polars
+    if writer is not None:
+        buf = io.BytesIO()
+        writer(buf)
+        return f"sim_{index}.parquet", buf.getvalue()
+
+    writer = getattr(entry, "to_parquet", None)          # pandas
+    if writer is not None:
+        buf = io.BytesIO()
+        writer(buf, index=False)
+        return f"sim_{index}.parquet", buf.getvalue()
+
+    path = Path(entry)
+    return path.name, path.read_bytes()
+
+
 def _build_config(
     run_metrics,
     run_impact,
@@ -66,12 +111,14 @@ def _build_config(
     plot_data,
     n_levels,
     l2_only,
+    **extra,
 ) -> dict:
     """The validation config object as the API expects it.
 
     Tri-state flags left as None are omitted rather than sent as null: the API
     reads absence as "use my tier's default", and an explicit null would not
-    do that.
+    do that. The same rule covers the 1.10.0 area flags and everything in
+    ``extra`` — naming nothing new leaves the job behaving exactly as before.
     """
     config = {
         "n_levels": n_levels,
@@ -84,6 +131,18 @@ def _build_config(
     ):
         if value is not None:
             config[flag] = value
+
+    for name in _AREA_FLAGS + _EXTRA_FIELDS:
+        value = extra.get(name)
+        if value is not None:
+            config[name] = value
+
+    unknown = set(extra) - set(_AREA_FLAGS) - set(_EXTRA_FIELDS)
+    if unknown:
+        raise ValueError(
+            f"Unknown validation option(s): {sorted(unknown)}. "
+            f"Valid: {sorted(_AREA_FLAGS + _EXTRA_FIELDS)}"
+        )
     return config
 
 
@@ -106,6 +165,17 @@ class ValidationResource:
         l2_only: bool = False,
         provider: str = None,
         exchange: str = None,
+        statistical: bool = None,
+        stylised_facts: bool = None,
+        impact: bool = None,
+        volume_correlation: bool = None,
+        fid: bool = None,
+        mind: bool = None,
+        lob: bool = None,
+        sample_period: str = None,
+        match_generated_sample: bool = None,
+        plots=None,
+        historical_output: bool = None,
     ) -> dict:
         """Submit a validation job.
 
@@ -155,6 +225,17 @@ class ValidationResource:
         config = _build_config(
             run_metrics, run_impact, run_inception_distances,
             run_stylised_facts, plot_data, n_levels, l2_only,
+            statistical=statistical,
+            stylised_facts=stylised_facts,
+            impact=impact,
+            volume_correlation=volume_correlation,
+            fid=fid,
+            mind=mind,
+            lob=lob,
+            sample_period=sample_period,
+            match_generated_sample=match_generated_sample,
+            plots=plots,
+            historical_output=historical_output,
         )
 
         payload = {
@@ -186,6 +267,17 @@ class ValidationResource:
         plot_data: bool = None,
         n_levels: int = 10,
         l2_only: bool = False,
+        statistical: bool = None,
+        stylised_facts: bool = None,
+        impact: bool = None,
+        volume_correlation: bool = None,
+        fid: bool = None,
+        mind: bool = None,
+        lob: bool = None,
+        sample_period: str = None,
+        match_generated_sample: bool = None,
+        plots=None,
+        historical_output: bool = None,
     ) -> dict:
         """Submit a validation job from simulation files you hold yourself.
 
@@ -227,15 +319,22 @@ class ValidationResource:
         config = _build_config(
             run_metrics, run_impact, run_inception_distances,
             run_stylised_facts, plot_data, n_levels, l2_only,
+            statistical=statistical,
+            stylised_facts=stylised_facts,
+            impact=impact,
+            volume_correlation=volume_correlation,
+            fid=fid,
+            mind=mind,
+            lob=lob,
+            sample_period=sample_period,
+            match_generated_sample=match_generated_sample,
+            plots=plots,
+            historical_output=historical_output,
         )
 
         files = []
-        for entry in sim_files:
-            if isinstance(entry, tuple):
-                filename, content = entry
-            else:
-                path = Path(entry)
-                filename, content = path.name, path.read_bytes()
+        for index, entry in enumerate(sim_files):
+            filename, content = _frame_to_parquet(entry, index)
             files.append(
                 ("sim_files", (filename, content, "application/octet-stream"))
             )
@@ -316,6 +415,17 @@ class ValidationResource:
         exchange: str = None,
         poll_interval: float = 3.0,
         timeout: float = 600.0,
+        statistical: bool = None,
+        stylised_facts: bool = None,
+        impact: bool = None,
+        volume_correlation: bool = None,
+        fid: bool = None,
+        mind: bool = None,
+        lob: bool = None,
+        sample_period: str = None,
+        match_generated_sample: bool = None,
+        plots=None,
+        historical_output: bool = None,
     ) -> dict:
         """Submit a validation job and block until it completes.
 
@@ -364,6 +474,17 @@ class ValidationResource:
             l2_only=l2_only,
             provider=provider,
             exchange=exchange,
+            statistical=statistical,
+            stylised_facts=stylised_facts,
+            impact=impact,
+            volume_correlation=volume_correlation,
+            fid=fid,
+            mind=mind,
+            lob=lob,
+            sample_period=sample_period,
+            match_generated_sample=match_generated_sample,
+            plots=plots,
+            historical_output=historical_output,
         )
         job_id = job["job_id"]
         print(f"Validation job submitted: {job_id}", file=sys.stderr)

{simudyne_pulse-0.7.0.dev2 → simudyne_pulse-0.7.0.dev4}/src/simudyne_pulse.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: simudyne-pulse
-Version: 0.7.0.dev2
+Version: 0.7.0.dev4
 Summary: Python SDK for the Simudyne Pulse synthetic market data API
 Author-email: Simudyne <support@simudyne.com>
 License-Expression: MIT

{simudyne_pulse-0.7.0.dev2 → simudyne_pulse-0.7.0.dev4}/tests/test_validation.py

@@ -156,3 +156,113 @@ class TestRunPipeline:
         )
         assert result["status"] == "completed"
         assert "spread" in result["distances"]
+
+
+class TestNewValidationOptions:
+    """pulse-check 1.10.0 areas, sampling and plot selection.
+
+    Everything here is opt-in: a job naming none of it must send exactly the
+    config it sent before, so existing callers are untouched.
+    """
+
+    BASE = dict(
+        run_metrics=None,
+        run_impact=None,
+        run_inception_distances=True,
+        run_stylised_facts=None,
+        plot_data=None,
+        n_levels=10,
+        l2_only=False,
+    )
+
+    def test_naming_nothing_new_is_unchanged(self):
+        from simudyne.resources.validation import _build_config
+
+        assert _build_config(**self.BASE) == {
+            "n_levels": 10,
+            "l2_only": False,
+            "run_fid": True,
+        }
+
+    @pytest.mark.parametrize(
+        "name,value",
+        [
+            ("statistical", False),
+            ("stylised_facts", True),
+            ("impact", False),
+            ("volume_correlation", True),
+            ("fid", False),
+            ("mind", True),
+            ("lob", True),
+            ("sample_period", "100ms"),
+            ("match_generated_sample", True),
+            ("historical_output", True),
+            ("plots", ["volume_correlation.levels"]),
+        ],
+    )
+    def test_option_is_forwarded(self, name, value):
+        from simudyne.resources.validation import _build_config
+
+        assert _build_config(**self.BASE, **{name: value})[name] == value
+
+    def test_unknown_option_is_rejected_by_name(self):
+        from simudyne.resources.validation import _build_config
+
+        with pytest.raises(ValueError, match="Unknown validation option"):
+            _build_config(**self.BASE, volume_corelation=True)
+
+    def test_none_is_omitted_not_sent_as_null(self):
+        """Absence means 'use my tier's default'; null would not."""
+        from simudyne.resources.validation import _build_config
+
+        assert "volume_correlation" not in _build_config(
+            **self.BASE, volume_correlation=None
+        )
+
+
+class TestSimulatedFrameShapes:
+    """run_upload takes a path, a (name, bytes) pair, or a DataFrame."""
+
+    def test_tuple_passes_through(self):
+        from simudyne.resources.validation import _frame_to_parquet
+
+        assert _frame_to_parquet(("a.parquet", b"raw"), 0) == ("a.parquet", b"raw")
+
+    def test_path(self, tmp_path):
+        import pandas as pd
+
+        from simudyne.resources.validation import _frame_to_parquet
+
+        p = tmp_path / "run.parquet"
+        pd.DataFrame({"a": [1, 2]}).to_parquet(p)
+        name, content = _frame_to_parquet(str(p), 0)
+        assert name == "run.parquet"
+        assert content[:4] == b"PAR1"
+
+    def test_pandas_frame(self):
+        import pandas as pd
+
+        from simudyne.resources.validation import _frame_to_parquet
+
+        name, content = _frame_to_parquet(pd.DataFrame({"a": [1, 2]}), 3)
+        assert name == "sim_3.parquet"
+        assert content[:4] == b"PAR1"
+
+    def test_polars_frame(self):
+        pl = pytest.importorskip("polars")
+
+        from simudyne.resources.validation import _frame_to_parquet
+
+        name, content = _frame_to_parquet(pl.DataFrame({"a": [1, 2]}), 1)
+        assert name == "sim_1.parquet"
+        assert content[:4] == b"PAR1"
+
+    def test_each_frame_gets_its_own_name(self):
+        """Distinct names, or the multipart upload collapses the runs."""
+        import pandas as pd
+
+        from simudyne.resources.validation import _frame_to_parquet
+
+        frames = [pd.DataFrame({"a": [i]}) for i in range(3)]
+        names = [_frame_to_parquet(f, i)[0] for i, f in enumerate(frames)]
+        assert len(set(names)) == 3