rahil-clm 0.1.1__tar.gz → 0.1.2__tar.gz

{rahil_clm-0.1.1 → rahil_clm-0.1.2}/PKG-INFO

@@ -1,10 +1,11 @@
 Metadata-Version: 2.4
 Name: rahil-clm
-Version: 0.1.1
+Version: 0.1.2
 Summary: Generate LHS samples and CLM crop PFT NetCDF parameter files for yield optimization.
 Author: Mohammad Uzair Rahil
 License: MIT
 Requires-Python: >=3.9
+Requires-Dist: matplotlib>=3.6
 Requires-Dist: netcdf4>=1.6
 Requires-Dist: numpy<2
 Requires-Dist: openpyxl>=3.1

{rahil_clm-0.1.1 → rahil_clm-0.1.2}/pyproject.toml

@@ -4,20 +4,22 @@ build-backend = "hatchling.build"
 
 [project]
 name = "rahil-clm"
-version = "0.1.1"
+version = "0.1.2"
 
 description = "Generate LHS samples and CLM crop PFT NetCDF parameter files for yield optimization."
 readme = "README.md"
 requires-python = ">=3.9"
 license = { text = "MIT" }
 authors = [{ name = "Mohammad Uzair Rahil" }]
+
 dependencies = [
   "numpy<2",
   "pandas>=1.5",
   "xarray>=2023.1",
   "netcdf4>=1.6",
   "requests>=2.31",
-  "openpyxl>=3.1"
+  "openpyxl>=3.1",
+  "matplotlib>=3.6"
 ]
 
 

rahil_clm-0.1.2/src/rahil/__init__.py

@@ -0,0 +1,5 @@
+from .clm import generate_lhs, GenerateLHSResult, check_distribution
+
+__all__ = ["generate_lhs", "GenerateLHSResult", "check_distribution"]
+__version__ = "0.1.2"
+

rahil_clm-0.1.2/src/rahil/clm/__init__.py

@@ -0,0 +1,4 @@
+from .core import generate_lhs, GenerateLHSResult
+from .plotting import check_distribution
+
+__all__ = ["generate_lhs", "GenerateLHSResult", "check_distribution"]

rahil_clm-0.1.2/src/rahil/clm/plotting.py

@@ -0,0 +1,244 @@
+from __future__ import annotations
+
+import os
+import math
+import glob
+import numpy as np
+import pandas as pd
+import matplotlib.pyplot as plt
+
+# Reuse your core download/cache logic + constants
+from .core import _ensure_inputs  # must exist in your core.py
+
+
+# ============================================================
+# Defaults (must match your LHS generator)
+# ============================================================
+PFTS = {"corn": 17, "soybean": 23, "wheat": 19}
+
+BOUND_COLS = {
+    "corn":    ("corn min", "corn max"),
+    "soybean": ("soybean min", "soybean max"),
+    "wheat":   ("wheat min", "wheat max"),
+}
+
+NS_PER_DAY = 86400.0 * 1e9
+
+
+# ============================================================
+# Helpers
+# ============================================================
+def _find_latest_param_list(workflow_dir: str) -> str:
+    candidates = glob.glob(os.path.join(workflow_dir, "*.param_list.txt"))
+    if not candidates:
+        raise FileNotFoundError(
+            f"No *.param_list.txt found in workflow dir: {workflow_dir}\n"
+            "Make sure you already ran rahil.generate_lhs(output_dir=...)."
+        )
+    candidates.sort(key=os.path.getmtime, reverse=True)
+    return candidates[0]
+
+
+def _normalize_columns(df: pd.DataFrame) -> pd.DataFrame:
+    df = df.copy()
+    df.columns = df.columns.astype(str).str.strip()
+    return df
+
+
+def _normalize_parameter_index(bounds_df: pd.DataFrame) -> pd.DataFrame:
+    bounds_df = _normalize_columns(bounds_df)
+
+    # normalize parameter column name -> "Parameters"
+    rename_map = {}
+    for c in bounds_df.columns:
+        if str(c).strip().lower() in ["parameter", "parameters", "param", "par"]:
+            rename_map[c] = "Parameters"
+    bounds_df = bounds_df.rename(columns=rename_map)
+
+    if "Parameters" not in bounds_df.columns:
+        raise KeyError(
+            "No parameter column found in bounds file.\n"
+            f"Available columns: {list(bounds_df.columns)}"
+        )
+
+    bounds_df["Parameters"] = bounds_df["Parameters"].astype(str).str.strip()
+    bounds_df = bounds_df.dropna(subset=["Parameters"]).set_index("Parameters")
+    bounds_df = bounds_df[~bounds_df.index.duplicated(keep="first")]
+    return bounds_df
+
+
+def _infer_param_names_from_sampled_columns(cols) -> list[str]:
+    params = set()
+    for c in cols:
+        c = str(c)
+        if "__pft" in c:
+            params.add(c.split("__pft")[0].strip())
+    return sorted(params)
+
+
+def _load_bounds_excel(bounds_xlsx_path: str, sheet_name=None) -> pd.DataFrame:
+    """
+    Load bounds from Excel. If sheet_name is None, try to find a sheet that contains
+    required columns; otherwise use the provided sheet.
+    """
+    xls = pd.ExcelFile(bounds_xlsx_path)
+
+    if sheet_name is not None:
+        b = pd.read_excel(bounds_xlsx_path, sheet_name=sheet_name)
+        return _normalize_parameter_index(b)
+
+    required = {
+        "Parameters",
+        "corn min", "corn max",
+        "soybean min", "soybean max",
+        "wheat min", "wheat max",
+    }
+
+    for sh in xls.sheet_names:
+        b = pd.read_excel(bounds_xlsx_path, sheet_name=sh)
+        b.columns = b.columns.str.strip()
+        b = b.rename(columns={"parameter": "Parameters", "Parameter": "Parameters", "PARAMETER": "Parameters"})
+        if required.issubset(set(b.columns)):
+            return _normalize_parameter_index(b)
+
+    # fallback: first sheet, but will raise meaningful error if missing columns
+    b = pd.read_excel(bounds_xlsx_path, sheet_name=xls.sheet_names[0])
+    return _normalize_parameter_index(b)
+
+
+# ============================================================
+# Public API
+# ============================================================
+def check_distribution(
+    output_dir: str = "outputs",
+    *,
+    sampled_file: str | None = None,
+    bounds_sheet=None,
+    outdir: str | None = None,
+    n_cols: int = 4,
+    panel_w: float = 3.2,
+    panel_h: float = 2.4,
+    show: bool = False,
+    dpi: int = 300,
+    cache_dir: str | None = None,
+) -> str:
+    """
+    Create distribution plots (one figure per crop) for whichever samples exist
+    in output_dir/workflow/*.param_list.txt.
+
+    Parameters
+    ----------
+    output_dir : str
+        Folder used in rahil.generate_lhs(output_dir=...).
+    sampled_file : str | None
+        Optional explicit path to *.param_list.txt. If None, automatically picks newest.
+    bounds_sheet : str|int|None
+        Optional bounds Excel sheet name (or index). If None, auto-detect.
+    outdir : str | None
+        Where to save plots. Default: output_dir/figs_distributions_by_crop
+    n_cols : int
+        Number of columns in subplot grid (article layout).
+    panel_w, panel_h : float
+        Size per subplot (inches).
+    show : bool
+        If True, plt.show() each figure. Default False (faster, cleaner).
+    dpi : int
+        Saved figure dpi.
+    cache_dir : str | None
+        Optional cache dir for bounds/base files. If None, core decides.
+
+    Returns
+    -------
+    str : directory path where figures were saved
+    """
+    # 1) locate sampled param list
+    workflow_dir = os.path.join(output_dir, "workflow")
+    if sampled_file is None:
+        sampled_file = _find_latest_param_list(workflow_dir)
+
+    # 2) ensure bounds file exists (download/cache from GitHub release)
+    bounds_xlsx_path, _, _ = _ensure_inputs(cache_dir=cache_dir)
+
+    # 3) read sampled + bounds
+    df = pd.read_csv(sampled_file, index_col=0)
+    df = _normalize_columns(df)
+
+    bounds = _load_bounds_excel(bounds_xlsx_path, sheet_name=bounds_sheet)
+    bounds.index = bounds.index.astype(str).str.strip()
+
+    sampled_params = _infer_param_names_from_sampled_columns(df.columns)
+    params_to_plot = [p for p in bounds.index if p in sampled_params]
+
+    # 4) output plot directory
+    if outdir is None:
+        outdir = os.path.join(output_dir, "figs_distributions_by_crop")
+    os.makedirs(outdir, exist_ok=True)
+
+    plt.rcParams.update({"axes.grid": True, "font.size": 9})
+
+    # 5) one figure per crop
+    for crop, pid in PFTS.items():
+        min_col, max_col = BOUND_COLS[crop]
+
+        n = len(params_to_plot)
+        if n == 0:
+            raise ValueError(
+                "No parameters found to plot.\n"
+                "Check that your sampled param_list columns match the bounds table parameter names."
+            )
+
+        ncols = min(n_cols, max(1, n))
+        nrows = int(math.ceil(n / ncols))
+        fig_w = panel_w * ncols
+        fig_h = panel_h * nrows
+
+        fig, axes = plt.subplots(nrows=nrows, ncols=ncols, figsize=(fig_w, fig_h))
+        axes = np.array(axes).reshape(-1)
+
+        fig.suptitle(
+            f"{crop.capitalize()} (PFT {pid}) — LHS sampled distributions",
+            fontsize=14, y=0.995
+        )
+
+        for i, param in enumerate(params_to_plot):
+            ax = axes[i]
+            col = f"{param}__pft{pid}_{crop}"
+
+            if col not in df.columns:
+                ax.set_title(f"{param} (missing)", fontsize=9)
+                ax.axis("off")
+                continue
+
+            vals = pd.to_numeric(df[col], errors="coerce").dropna().to_numpy()
+
+            vmin = float(bounds.loc[param, min_col])
+            vmax = float(bounds.loc[param, max_col])
+
+            # mxmat bounds may be ns in excel; values are days in sampled list
+            if param == "mxmat":
+                if abs(vmin) > 1e6 or abs(vmax) > 1e6:
+                    vmin /= NS_PER_DAY
+                    vmax /= NS_PER_DAY
+                vals = np.rint(vals).astype(int)
+
+            ax.hist(vals, bins=20, edgecolor="k", alpha=0.75)
+            ax.axvline(vmin, color="red", linestyle="--", linewidth=1)
+            ax.axvline(vmax, color="red", linestyle="--", linewidth=1)
+
+            ax.set_title(param, fontsize=10)
+            ax.set_ylabel("count", fontsize=8)
+            ax.set_xlabel("value" + (" (days)" if param == "mxmat" else ""), fontsize=8)
+            ax.tick_params(axis="both", labelsize=8)
+
+        for j in range(n, len(axes)):
+            axes[j].axis("off")
+
+        plt.tight_layout(rect=[0, 0, 1, 0.97])
+        fig.savefig(os.path.join(outdir, f"LHS_distributions_{crop}.png"),
+                    dpi=dpi, bbox_inches="tight")
+
+        if show:
+            plt.show()
+        plt.close(fig)
+
+    return outdir

rahil_clm-0.1.1/src/rahil/__init__.py

@@ -1,4 +0,0 @@
-from .clm import generate_lhs, GenerateLHSResult
-
-__all__ = ["generate_lhs", "GenerateLHSResult"]
-__version__ = "0.1.1"

rahil_clm-0.1.1/src/rahil/clm/__init__.py

@@ -1,3 +0,0 @@
-from .core import generate_lhs, GenerateLHSResult
-
-__all__ = ["generate_lhs", "GenerateLHSResult"]