Features/viewer

2016-04-25_diagnostics_summary.txt

@@ -0,0 +1,32 @@
+Diagnostics summary
+===================
+
+Total N (FIMS, cm^-3): 847.085
+Total N (pop,  cm^-3): 847.085
+Ratio (pop/FIMS):      1
+
+AMS mass fractions (observed, normalized over available keys):
+  SO4: 0.3376
+  NO3: 0.0352
+  OC: 0.5160
+  NH4: 0.1112
+
+AMS mass fractions (reconstructed, normalized over available keys):
+  SO4: 0.5251
+  NO3: 0.0085
+  OC: 0.1477
+  NH4: 0.3187
+
+miniSPLAT number fractions (observed, normalized):
+  AS: 0.6115
+  BB: 0.2991
+  BC: 0.0409
+  IEPOX: 0.0414
+  OIN: 0.0072
+
+miniSPLAT number fractions (reconstructed / builder metadata, normalized):
+  AS: 0.6115
+  BB: 0.2991
+  BC: 0.0409
+  IEPOX: 0.0414
+  OIN: 0.0072

README.pypi.md

@@ -51,6 +51,17 @@ print(pop)
 
 More examples are available in the project repository.
 
+## Interactive viewer
+
+An experimental Streamlit UI lets you build populations through the factory registries and render existing visualization builders.
+
+```bash
+pip install -e .
+streamlit run scripts/launch_viewer.py
+```
+
+The viewer source lives under `viewer/`. The sidebar lists every registered population and plot type; choose any combination, adjust the metadata-driven controls, and the figure plus diagnostics render inline.
+
 ## Contributing
 
 `part2pop` is designed so that **all extensibility happens through factories**.

diagnostics_summary.txt

@@ -0,0 +1,44 @@
+Diagnostics summary
+===================
+
+Fitted size distribution (Dpg_nm, sigma):
+  Mode 0: 73.9811, 1.682
+  Mode 1: 207.28, 1.357
+
+Optimized fraction of particles in each mode:
+  OC: 0.209 0.791 
+  SO4: 0.569 0.431 
+  NO3: 0.853 0.147 
+  IEPOX_SOA: 0.239 0.761 
+
+Total N (FIMS, cm^-3): 1582.01
+Total N (pop,  cm^-3): 1707.48
+Ratio (pop/FIMS):      1.07931
+
+AMS mass fractions (observed, normalized over available keys):
+  OC: 0.4764
+  NO3: 0.0670
+  SO4: 0.3183
+  NH4: 0.1383
+
+AMS mass fractions (reconstructed, normalized over available keys):
+  OC: 0.5267
+  NO3: 0.0481
+  SO4: 0.3226
+  NH4: 0.1026
+
+miniSPLAT number fractions (observed, normalized):
+  BC: 0.0108
+  OIN: 0.0042
+  OC: 0.4683
+  SO4: 0.3496
+  NO3: 0.1454
+  IEPOX_SOA: 0.0217
+
+miniSPLAT number fractions (reconstructed, normalized):
+  BC: 0.0108
+  OIN: 0.0042
+  OC: 0.4683
+  SO4: 0.3496
+  NO3: 0.1454
+  IEPOX_SOA: 0.0217

environment.yml

@@ -0,0 +1,16 @@
+name: part2pop
+channels:
+  - conda-forge
+dependencies:
+  - python>=3.10
+  - numpy
+  - scipy
+  - matplotlib
+  - netCDF4
+  - importlib-resources
+  - pymiescatt
+  - tqdm
+  - streamlit
+  - pip
+  - pip:
+    - -r requirements.txt

examples/example_data/hiscale_population_config.json

@@ -0,0 +1,100 @@
+{
+  "type": "hiscale_observations",
+  "aimms_file": "/Users/fier887/Library/CloudStorage/OneDrive-PNNL/Code/multipart_archived2/multipart_archived/separate_tools/datasets/HISCALE_data_0425/AIMMS20_G1_20160425155810_R2_HISCALE020h.txt",
+  "splat_file": "/Users/fier887/Library/CloudStorage/OneDrive-PNNL/Code/multipart_archived2/multipart_archived/separate_tools/datasets/HISCALE_data_0425/Splat_Composition_25-Apr-2016.txt",
+  "ams_file": "/Users/fier887/Library/CloudStorage/OneDrive-PNNL/Code/multipart_archived2/multipart_archived/separate_tools/datasets/HISCALE_data_0425/HiScaleAMS_G1_20160425_R0.txt",
+  "fims_file": "/Users/fier887/Library/CloudStorage/OneDrive-PNNL/Code/multipart_archived2/multipart_archived/separate_tools/datasets/HISCALE_data_0425/FIMS_G1_20160425_R1_HISCALE_001s.txt",
+  "fims_bins_file": "/Users/fier887/Library/CloudStorage/OneDrive-PNNL/Code/multipart_archived2/multipart_archived/separate_tools/datasets/HISCALE_data_0425/HISCALE_FIMS_bins_R1.txt",
+  "z": 1000.0,
+  "dz": 100.0,
+  "splat_species": {
+    "BC": [
+      "soot"
+    ],
+    "OIN": [
+      "Dust"
+    ],
+    "OC": [
+      "org28",
+      "org30_43",
+      "BB_SOA",
+      "org_amines",
+      "BB",
+      "pyridine"
+    ],
+    "SO4": [
+      "sulfate_nitrate_org"
+    ],
+    "NO3": [
+      "nitrate_amine_org"
+    ],
+    "IEPOX_SOA": [
+      "IEPOX_SOA"
+    ]
+  },
+  "mass_thresholds": {
+    "IEPOX_SOA": [
+      [
+        0.3,
+        0.5,
+        0.1
+      ],
+      [
+        "IEPOX_OS",
+        "tetrol",
+        "tetrol_olig",
+        "IEPOX_OH_SOA"
+      ]
+    ],
+    "SO4": [
+      [
+        0.5,
+        0.7,
+        0.1
+      ],
+      [
+        "SO4"
+      ]
+    ],
+    "NO3": [
+      [
+        0.5,
+        0.7,
+        0.1
+      ],
+      [
+        "NO3"
+      ]
+    ],
+    "OC": [
+      [
+        0.5,
+        0.7,
+        0.1
+      ],
+      [
+        "OC"
+      ]
+    ],
+    "BC": [
+      [
+        0.5,
+        0.7,
+        0.1
+      ],
+      [
+        "BC"
+      ]
+    ],
+    "OIN": [
+      [
+        0.5,
+        0.7,
+        0.1
+      ],
+      [
+        "OIN"
+      ]
+    ]
+  }
+}

launch_viewer.py

@@ -0,0 +1,5 @@
+from viewer import run_viewer
+
+
+if __name__ == "__main__":
+    run_viewer()

requirements.txt

@@ -0,0 +1,11 @@
+numpy
+scipy
+matplotlib
+netCDF4
+importlib-resources
+PyMieScatt
+pyBCabs
+tqdm
+streamlit
+pytest>=7
+pytest-cov>=4

scripts/launch_viewer.py

@@ -0,0 +1,19 @@
+"""Entry point for running the Streamlit viewer in a source checkout."""
+
+from __future__ import annotations
+
+from pathlib import Path
+import sys
+
+REPO_ROOT = Path(__file__).resolve().parents[1]
+SRC_DIR = REPO_ROOT / "src"
+if SRC_DIR.is_dir() and str(SRC_DIR) not in sys.path:
+    sys.path.insert(0, str(SRC_DIR))
+if str(REPO_ROOT) not in sys.path:
+    sys.path.insert(0, str(REPO_ROOT))
+
+from viewer import run_viewer
+
+
+if __name__ == "__main__":
+    run_viewer()

src/part2pop/analysis/defaults.py

@@ -80,7 +80,12 @@
         "T": 298.15,
     },
 
-    # relative-humidity axis
+    # temperature / humidity axes
+    "T_grid": {
+        "T_grid": np.asarray([298.15]),
+        "T": 298.15,
+        "T_units": "K",
+    },
     "rh_grid": {"rh_grid": np.asarray([0.0])},
 
     # other/legacy entries (keep for compatibility)
@@ -110,4 +115,3 @@ def get_defaults_for_variable(name: str) -> Dict[str, Any]:
 def all_defaults() -> Dict[str, Dict[str, Any]]:
     """Return a copy of the whole defaults mapping (diagnostic)."""
     return {k: dict(v) for k, v in _DEFAULTS_BY_VAR.items()}
-

src/part2pop/analysis/distributions.py

@@ -2,6 +2,16 @@
 
 import numpy as np
 
+_TRAPEZOID_FUNC = getattr(np, "trapezoid", None)
+_TRAPZ_FUNC = getattr(np, "trapz", None)
+
+if _TRAPZ_FUNC is None and _TRAPEZOID_FUNC is not None:
+    _TRAPZ_FUNC = _TRAPEZOID_FUNC
+    np.trapz = _TRAPZ_FUNC
+
+if _TRAPEZOID_FUNC is None and _TRAPZ_FUNC is not None:
+    _TRAPEZOID_FUNC = _TRAPZ_FUNC
+    np.trapezoid = _TRAPEZOID_FUNC
 
 def _trapezoid_integrate(y, x=None, dx=1.0, axis=-1):
     """
@@ -11,12 +21,17 @@ def _trapezoid_integrate(y, x=None, dx=1.0, axis=-1):
     releases only expose `np.trapz`.  This helper always calls whichever
     function exists so the rest of the module can rely on a single name.
     """
-    func = getattr(np, "trapezoid", None)
-    if func is None:
-        func = getattr(np, "trapz", None)
-    if func is None:
-        raise AttributeError("NumPy installation lacks trapezoid/trapz integrators")
-    return func(y, x=x, dx=dx, axis=axis)
+    if _TRAPEZOID_FUNC is not None:
+        return _TRAPEZOID_FUNC(y, x=x, dx=dx, axis=axis)
+    if _TRAPZ_FUNC is not None:
+        return _TRAPZ_FUNC(y, x=x, dx=dx, axis=axis)
+    raise AttributeError("NumPy installation lacks trapezoid/trapz integrators")
+    # func = getattr(np, "trapezoid", None)
+    # if func is None:
+    #     func = getattr(np, "trapz", None)
+    # if func is None:
+    #     raise AttributeError("NumPy installation lacks trapezoid/trapz integrators")
+    # return func(y, x=x, dx=dx, axis=axis)
 
 try:
     from scipy.interpolate import PchipInterpolator as _PCHIP

src/part2pop/analysis/particle/factory/registry.py

@@ -110,13 +110,23 @@ def describe_particle_variable(name: str):
     if not meta:
         raise UnknownParticleVariableError(name, suggestions=None)
 
+    axis_names = meta.axis_names
+    if isinstance(axis_names, str):
+        axis_names = [axis_names]
+    else:
+        axis_names = list(axis_names)
+
+    units = getattr(meta, "units", None)
+    if isinstance(units, dict):
+        units = dict(units)
+
     return {
         "name": meta.name,
         "value_key": meta.name,
-        "axis_keys": list(meta.axis_names),
+        "axis_keys": list(axis_names),
+        "axis_names": list(axis_names),
         "description": meta.description,
         "aliases": list(meta.aliases),
         "defaults": dict(meta.default_cfg),
-        "units": dict(meta.units) if getattr(meta, "units", None) else None,
+        "units": units,
     }
-

src/part2pop/analysis/population/factory/dNdlnD.py

@@ -59,7 +59,7 @@ def compute(self, population, as_dict: bool = False):
         - The variable is *always* defined w.r.t. ln(D), i.e. measure="ln".
         """
         cfg = self.cfg
-        method = cfg.get("method", "hist")
+        method = cfg.get("method", "kde")
         measure = "ln"  # this variable is per dlnD by definition
 
         # ------------------------------------------------------------------
@@ -201,7 +201,7 @@ def build(cfg=None) -> DNdlnDVar:
     Factory function used by the analysis population registry.
 
     Config keys (common ones):
-      - "method": "hist" (default) or "kde"
+      - "method": "kde" (default) or "hist"
       - "wetsize": bool (True = use wet diameters, False = dry)
       - "N_bins": int
       - "D_min", "D_max": floats in meters

src/part2pop/population/factory/hiscale_observations.py

@@ -288,12 +288,12 @@ def _read_fims_bins_file(fims_bins_file: str, expected_bins: int) -> Tuple[np.nd
     Parse a bins file and return (Dp_lowers_nm, Dp_uppers_nm) with length expected_bins.
     Works for:
       - 2-column tables (lower upper)
+      - 3-column tables (min, mean, max) where we use first/last columns
       - numeric streams that are interleaved lo,hi,lo,hi...
       - numeric streams that are concatenated all-lo then all-hi
     """
     lines = _read_lines(fims_bins_file)
 
-    # Try per-line 2-col parse first
     lo_list: List[float] = []
     hi_list: List[float] = []
     for line in lines:
@@ -303,7 +303,10 @@ def _read_fims_bins_file(fims_bins_file: str, expected_bins: int) -> Tuple[np.nd
         vals = re.findall(r"[-+]?\d*\.?\d+(?:[eE][-+]?\d+)?", s)
         if len(vals) >= 2:
             lo_list.append(float(vals[0]))
-            hi_list.append(float(vals[1]))
+            hi_list.append(float(vals[-1]))
+        if len(vals) >= 3 and vals[0] != vals[-1]:
+            lo_list[-1] = float(vals[0])
+            hi_list[-1] = float(vals[-1])
     if len(lo_list) == expected_bins and len(hi_list) == expected_bins:
         lo = np.array(lo_list, dtype="float64")
         hi = np.array(hi_list, dtype="float64")