process-improve 1.75.2__tar.gz → 1.82.0__tar.gz

{process_improve-1.75.2 → process_improve-1.82.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: process-improve
-Version: 1.75.2
+Version: 1.82.0
 Summary: Designed Experiments; Latent Variables (PCA, PLS, multivariate methods with missing data); Process Monitoring; Batch data analysis.
 Keywords: Designed Experiments,Latent Variables,PCA,PLS,Multivariate Data Analysis,Batch data analysis
 Author: Kevin Dunn
@@ -28,6 +28,7 @@ Requires-Dist: matplotlib>=3.10.8 ; extra == 'all'
 Requires-Dist: mcp>=2.0 ; extra == 'all'
 Requires-Dist: numba>=0.63.1 ; extra == 'all'
 Requires-Dist: openpyxl>=3.1.5 ; extra == 'all'
+Requires-Dist: osqp>=1.0 ; extra == 'all'
 Requires-Dist: plotly>=6.5.2 ; extra == 'all'
 Requires-Dist: pulp>=2.8 ; extra == 'all'
 Requires-Dist: pydoe3>=1.0 ; extra == 'all'
@@ -36,6 +37,7 @@ Requires-Dist: scikit-image>=0.25.2 ; extra == 'all'
 Requires-Dist: seaborn>=0.13.2 ; extra == 'all'
 Requires-Dist: openpyxl>=3.1.5 ; extra == 'batch'
 Requires-Dist: scikit-image>=0.25.2 ; extra == 'batch'
+Requires-Dist: osqp>=1.0 ; extra == 'control'
 Requires-Dist: coverage>=7.13.1 ; extra == 'dev'
 Requires-Dist: hypothesis>=6.100 ; extra == 'dev'
 Requires-Dist: matplotlib-stubs>=0.3.11 ; extra == 'dev'
@@ -63,11 +65,12 @@ Requires-Dist: plotly>=6.5.2 ; extra == 'plotting'
 Requires-Dist: ridgeplot>=0.5.0 ; extra == 'plotting'
 Requires-Dist: seaborn>=0.13.2 ; extra == 'plotting'
 Requires-Python: >=3.10
-Project-URL: Homepage, https://github.com/kgdunn/process_improve
-Project-URL: Repository, https://github.com/kgdunn/process_improve
-Project-URL: Issues, https://github.com/kgdunn/process_improve/issues
+Project-URL: Homepage, https://github.com/kgdunn/process-improve
+Project-URL: Repository, https://github.com/kgdunn/process-improve
+Project-URL: Issues, https://github.com/kgdunn/process-improve/issues
 Provides-Extra: all
 Provides-Extra: batch
+Provides-Extra: control
 Provides-Extra: dev
 Provides-Extra: expt
 Provides-Extra: fast
@@ -268,8 +271,8 @@ result.y_hat                      # point predictions
 result.spe                        # squared prediction error
 result.hotellings_t2              # Hotelling's T² for new observations
 
-# Cross-validated component selection
-cv_select = PLS.select_n_components(X_s, Y_s, max_components=6)
+# Cross-validated component selection: raw blocks in, each fold scales itself
+cv_select = PLS.select_n_components(X, Y, max_components=6)
 print(cv_select.n_components)     # recommended number of components
 print(cv_select.rmsecv)           # RMSECV per component count
 

{process_improve-1.75.2 → process_improve-1.82.0}/README.md

@@ -190,8 +190,8 @@ result.y_hat                      # point predictions
 result.spe                        # squared prediction error
 result.hotellings_t2              # Hotelling's T² for new observations
 
-# Cross-validated component selection
-cv_select = PLS.select_n_components(X_s, Y_s, max_components=6)
+# Cross-validated component selection: raw blocks in, each fold scales itself
+cv_select = PLS.select_n_components(X, Y, max_components=6)
 print(cv_select.n_components)     # recommended number of components
 print(cv_select.rmsecv)           # RMSECV per component count
 

{process_improve-1.75.2 → process_improve-1.82.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "process-improve"
-version = "1.75.2"
+version = "1.82.0"
 description = "Designed Experiments; Latent Variables (PCA, PLS, multivariate methods with missing data); Process Monitoring; Batch data analysis."
 readme = "README.md"
 license = "MIT"
@@ -55,11 +55,13 @@ batch = [
 mcp = ["mcp>=2.0"]
 fast = ["numba>=0.63.1"]
 ilp = ["pulp>=2.8"]
+control = ["osqp>=1.0"]
 all = [
     "matplotlib>=3.10.8",
     "mcp>=2.0",
     "numba>=0.63.1",
     "openpyxl>=3.1.5",
+    "osqp>=1.0",
     "plotly>=6.5.2",
     "pulp>=2.8",
     "pyDOE3>=1.0",
@@ -92,9 +94,9 @@ dev = [
 process-improve-mcp = "process_improve.mcp_server:main"
 
 [project.urls]
-Homepage = "https://github.com/kgdunn/process_improve"
-Repository = "https://github.com/kgdunn/process_improve"
-Issues = "https://github.com/kgdunn/process_improve/issues"
+Homepage = "https://github.com/kgdunn/process-improve"
+Repository = "https://github.com/kgdunn/process-improve"
+Issues = "https://github.com/kgdunn/process-improve/issues"
 
 [build-system]
 requires = ["uv_build>=0.7.2,<0.13"]
@@ -128,6 +130,7 @@ override-dependencies = [
     "plotly>=6.5.2",
     "numba>=0.63.1",
 ]
+constraint-dependencies = ["scipy!=1.18.1"]
 
 [tool.mypy]
 ignore_missing_imports = true

{process_improve-1.75.2 → process_improve-1.82.0}/pyproject.toml.orig

@@ -1,6 +1,6 @@
 [project]
 name = "process-improve"
-version = "1.75.2"
+version = "1.82.0"
 description = 'Designed Experiments; Latent Variables (PCA, PLS, multivariate methods with missing data); Process Monitoring; Batch data analysis.'
 readme = "README.md"
 license = "MIT"
@@ -81,6 +81,10 @@ fast = ["numba>=0.63.1"]
 # CBC solver in its wheels, so no system package is required.
 ilp = ["pulp>=2.8"]
 
+# Mid-course correction QP solver (batch.control). osqp ships prebuilt wheels
+# for every platform CI tests, so no system package is required.
+control = ["osqp>=1.0"]
+
 # NOTE: ``pyoptex`` (coordinate-exchange I-/A-optimal and split-plot designs in
 # ``experiments.designs_optimal``) is intentionally *not* a declared extra. Its
 # latest release (1.2.1) pins ``plotly~=5.24`` (i.e. < 6), which conflicts with
@@ -96,6 +100,7 @@ all = [
     "mcp>=2.0",
     "numba>=0.63.1",
     "openpyxl>=3.1.5",
+    "osqp>=1.0",
     "plotly>=6.5.2",
     "pulp>=2.8",
     "pyDOE3>=1.0",
@@ -129,9 +134,9 @@ dev = [
 process-improve-mcp = "process_improve.mcp_server:main"
 
 [project.urls]
-Homepage = "https://github.com/kgdunn/process_improve"
-Repository = "https://github.com/kgdunn/process_improve"
-Issues = "https://github.com/kgdunn/process_improve/issues"
+Homepage = "https://github.com/kgdunn/process-improve"
+Repository = "https://github.com/kgdunn/process-improve"
+Issues = "https://github.com/kgdunn/process-improve/issues"
 
 
 [build-system]
@@ -178,6 +183,14 @@ override-dependencies = [
     "plotly>=6.5.2",
     "numba>=0.63.1",
 ]
+# scipy 1.18.1 was published without CPython 3.13 wheels (cp312 only), and CI
+# re-resolves on a stale lockfile whenever this file changes, so a fresh
+# resolution on Python 3.13 picks it and fails to install. Exclude that one
+# release; drop this constraint once scipy uploads cp313 wheels for it or
+# publishes a newer release.
+constraint-dependencies = [
+    "scipy!=1.18.1",
+]
 
 
 [tool.mypy]

{process_improve-1.75.2 → process_improve-1.82.0}/src/process_improve/_extras.py

@@ -10,6 +10,7 @@ optional dependencies behind extras::
     pip install 'process-improve[mcp]'         # mcp
     pip install 'process-improve[fast]'        # numba (JIT)
     pip install 'process-improve[ilp]'         # pulp (OMARS design generator)
+    pip install 'process-improve[control]'     # osqp (mid-course correction QP)
     pip install 'process-improve[all]'         # everything above
 
 Modules that need an optional dependency import it inside a

{process_improve-1.75.2 → process_improve-1.82.0}/src/process_improve/_linalg.py

@@ -70,7 +70,14 @@ def safe_inverse(
     if is_singular(a, cond_limit=cond_limit):
         raise np.linalg.LinAlgError(
             f"{what} is singular or ill-conditioned (condition number exceeds "
-            f"{cond_limit:.2e}); cannot invert reliably. Reduce the number of "
-            "components/terms or remove collinear columns."
+            f"{cond_limit:.2e}); cannot invert reliably. Likely causes, in the "
+            "order worth checking: exactly or near-exactly collinear columns "
+            "(a duplicated variable, a column that is a sum of others, a "
+            "one-hot set that includes every level); a constant or all-missing "
+            "column; a component count that has exhausted the variance left in "
+            "the deflated block; or values spanning so many orders of magnitude "
+            "that the cross-product loses precision (centre and scale first). "
+            "Reduce the number of components/terms or remove the offending "
+            "columns."
         )
     return np.linalg.inv(a)

process_improve-1.82.0/src/process_improve/_remote_data.py

@@ -0,0 +1,147 @@
+# (c) Kevin Dunn, 2010-2026. MIT License. Based on own private work over the years.
+"""Bounded download of the sample datasets that are not bundled with the package.
+
+The remote sample datasets (the `openmv.net <https://openmv.net>`_ files behind
+:mod:`process_improve.experiments.datasets` and
+:mod:`process_improve.batch.datasets`) are fetched on demand. Every fetch goes
+through :func:`fetch_remote_bytes`, so one place enforces the contract:
+
+- the URL is a fixed ``https`` location chosen by the library, or a ``file://``
+  path a caller passes explicitly to read a local copy; it is never free text
+  from an untrusted source;
+- the download is bounded by ``settings.dataset_fetch_timeout`` (30 s by
+  default; ``PROCESS_IMPROVE_DATASET_FETCH_TIMEOUT`` overrides it), so a
+  black-holing host raises instead of hanging the caller indefinitely (#508);
+- network failures, timeouts and parse failures surface as one clear
+  ``RuntimeError`` naming the URL, rather than a lower-level error.
+
+The content is trusted only as far as the remote host is.
+"""
+
+from __future__ import annotations
+
+import io
+import urllib.request
+import zipfile
+
+import pandas as pd
+
+from process_improve._extras import require_extra
+from process_improve.config import settings
+
+
+def _download_error(url: str, exc: Exception) -> RuntimeError:
+    """Return the one ``RuntimeError`` every failure mode maps onto."""
+    return RuntimeError(
+        f"Could not download the sample dataset from {url!r}: {exc}. "
+        "Check your network connection; this dataset is fetched from a remote host."
+    )
+
+
+def fetch_remote_bytes(url: str, timeout: float | None = None) -> bytes:
+    """Download ``url`` and return the raw payload.
+
+    Parameters
+    ----------
+    url : str
+        The fixed ``https`` URL of the file to fetch, or a ``file://`` URL
+        for a local copy.
+    timeout : float, optional
+        Seconds to wait before giving up. Defaults to
+        ``settings.dataset_fetch_timeout``.
+
+    Returns
+    -------
+    bytes
+        The downloaded content.
+
+    Raises
+    ------
+    RuntimeError
+        On any network failure or timeout (``TimeoutError`` and
+        ``urllib.error.URLError`` are both ``OSError`` subclasses), with the
+        URL in the message.
+    """
+    if timeout is None:
+        timeout = settings.dataset_fetch_timeout
+    try:
+        with urllib.request.urlopen(url, timeout=timeout) as response:  # noqa: S310 - fixed https or file URLs
+            return response.read()
+    except OSError as exc:
+        raise _download_error(url, exc) from exc
+
+
+def read_remote_csv(url: str, timeout: float | None = None) -> pd.DataFrame:
+    """Fetch a CSV file with :func:`fetch_remote_bytes` and parse it.
+
+    Parameters
+    ----------
+    url : str
+        The fixed ``https`` (or ``file://``) URL of the CSV file.
+    timeout : float, optional
+        Seconds to wait before giving up. Defaults to
+        ``settings.dataset_fetch_timeout``.
+
+    Returns
+    -------
+    pd.DataFrame
+        The parsed table.
+
+    Raises
+    ------
+    RuntimeError
+        On a download failure or when the payload is not a readable CSV.
+    """
+    payload = fetch_remote_bytes(url, timeout=timeout)
+    try:
+        return pd.read_csv(io.BytesIO(payload))
+    except ValueError as exc:
+        raise _download_error(url, exc) from exc
+
+
+def read_remote_excel(
+    url: str,
+    timeout: float | None = None,
+    *,
+    sheet_name: str | list[str] | None = None,
+) -> dict[str, pd.DataFrame]:
+    """Fetch an Excel workbook with :func:`fetch_remote_bytes` and parse its sheets.
+
+    Reading ``.xlsx`` files needs ``openpyxl``, which ships with the ``batch``
+    extra; a missing install raises an ``ImportError`` that names the extra.
+
+    Parameters
+    ----------
+    url : str
+        The fixed ``https`` (or ``file://``) URL of the workbook.
+    timeout : float, optional
+        Seconds to wait before giving up. Defaults to
+        ``settings.dataset_fetch_timeout``.
+    sheet_name : str or list of str, optional
+        Sheets to read. ``None`` (default) reads every sheet.
+
+    Returns
+    -------
+    dict[str, pd.DataFrame]
+        One table per sheet, keyed by sheet name, also when a single sheet
+        name was requested.
+
+    Raises
+    ------
+    RuntimeError
+        On a download failure or when the payload is not a readable workbook.
+    ImportError
+        When ``openpyxl`` is not installed.
+    """
+    try:
+        import openpyxl  # noqa: F401, PLC0415 - probed before downloading so the error is immediate
+    except ImportError as exc:
+        raise require_extra("openpyxl", "batch") from exc
+    payload = fetch_remote_bytes(url, timeout=timeout)
+    try:
+        sheets = pd.read_excel(io.BytesIO(payload), sheet_name=sheet_name, engine="openpyxl")
+    except (ValueError, zipfile.BadZipFile) as exc:
+        raise _download_error(url, exc) from exc
+    if isinstance(sheets, pd.DataFrame):
+        return {str(sheet_name): sheets}
+    return {str(name): frame for name, frame in sheets.items()}

{process_improve-1.75.2 → process_improve-1.82.0}/src/process_improve/batch/__init__.py

@@ -1,13 +1,32 @@
-"""Batch process data analysis: alignment, feature extraction, and visualization."""
+"""Batch process data analysis: alignment, feature extraction, modelling, and visualization."""
 
+from process_improve.batch._batch_monitor import BatchMonitor
+from process_improve.batch._batch_pca import BatchPCA
+from process_improve.batch._batch_plots import (
+    contribution_at_time_plot,
+    online_monitoring_plot,
+    time_varying_loading_plot,
+    unfolded_contribution_plot,
+)
+from process_improve.batch._batch_pls import BatchPLS
+from process_improve.batch.control import MidCourseCorrector, midcourse_correction
 from process_improve.batch.data_input import (
     check_valid_batch_dict,
     dict_to_melted,
     dict_to_wide,
     melted_to_dict,
     melted_to_wide,
+    wide_to_dict,
     wide_to_melted,
 )
+from process_improve.batch.datasets import (
+    load_batch_fake_data,
+    load_dryer,
+    load_dupont,
+    load_fmc,
+    load_nylon,
+    load_sbr,
+)
 from process_improve.batch.features import (
     cross,
     f_agemax,
@@ -36,10 +55,16 @@ from process_improve.batch.preprocessing import (
 )
 
 __all__ = [
+    # Modelling and monitoring
+    "BatchMonitor",
+    "BatchPCA",
+    "BatchPLS",
+    "MidCourseCorrector",
     # Preprocessing/alignment
     "batch_dtw",
     # Data input/conversion
     "check_valid_batch_dict",
+    "contribution_at_time_plot",
     "cross",
     "determine_scaling",
     "dict_to_melted",
@@ -63,8 +88,20 @@ __all__ = [
     "f_std",
     "f_sum",
     "find_reference_batch",
+    # Bundled datasets
+    "load_batch_fake_data",
+    "load_dryer",
+    "load_dupont",
+    "load_fmc",
+    "load_nylon",
+    "load_sbr",
     "melted_to_dict",
     "melted_to_wide",
+    "midcourse_correction",
+    "online_monitoring_plot",
     "resample_to_reference",
+    "time_varying_loading_plot",
+    "unfolded_contribution_plot",
+    "wide_to_dict",
     "wide_to_melted",
 ]