Merge pull request #192 from GEECS-BELLA/Make-ScanPaths-attribute-of-ScanData-

Cleaner implementation of ScanData with additional simple visualization utilities

Author: skbarber

GEECS-Data-Utils/geecs_data_utils/__init__.py

@@ -13,7 +13,7 @@
 Examples
 --------
 >>> from geecs_data_utils import ScanData, ScanPaths
->>> scan_data = ScanData(year=2024, month=1, day=15, scan_number=42)
+>>> scan_paths = ScanData(year=2024, month=1, day=15, scan_number=42)
 """
 
 from geecs_data_utils.scan_data import ScanData

GEECS-Data-Utils/geecs_data_utils/plotting_utils.py

@@ -0,0 +1,296 @@
+"""
+Minimal plotting utilities for binned scan data.
+
+This module provides helper functions to plot aggregated (binned) scalar data
+from scans using a frozen schema. It supports both single and multi-series
+errorbar plots with optional asymmetric errors and index-based x-axes.
+
+Functions
+---------
+_get_center_and_err
+    Extract center values and optional asymmetric errors for a column.
+_index_to_numeric
+    Convert an index to numeric values, handling IntervalIndex specially.
+plot_binned
+    Plot a single y-series versus x from binned data.
+plot_binned_multi
+    Overlay multiple y-series versus the same x.
+"""
+
+from __future__ import annotations
+from typing import Optional, Sequence, Tuple
+import numpy as np
+import pandas as pd
+import matplotlib.pyplot as plt
+
+
+def _get_center_and_err(
+    binned: pd.DataFrame, col: str
+) -> Tuple[np.ndarray, Optional[np.ndarray]]:
+    """
+    Extract center values and optional errors for a column.
+
+    This function enforces a frozen schema:
+    - For MultiIndex columns, expects subcolumns 'center', and optionally
+      'err_low' and 'err_high'.
+    - For flat columns, returns values directly without errors.
+
+    Parameters
+    ----------
+    binned : pandas.DataFrame
+        DataFrame of binned values. May have a MultiIndex or flat columns.
+    col : str
+        Column name to extract values for.
+
+    Returns
+    -------
+    y : numpy.ndarray
+        Center values for the requested column.
+    yerr : numpy.ndarray or None
+        Asymmetric errors with shape (2, N) if both 'err_low' and 'err_high'
+        are present. Otherwise None.
+
+    Raises
+    ------
+    KeyError
+        If the requested column does not exist in the expected schema.
+    """
+    if isinstance(binned.columns, pd.MultiIndex):
+        cols = binned.columns
+        if (col, "center") in cols:
+            y = binned[(col, "center")].to_numpy()
+            if (col, "err_low") in cols and (col, "err_high") in cols:
+                yerr = np.vstack(
+                    [
+                        binned[(col, "err_low")].to_numpy(),
+                        binned[(col, "err_high")].to_numpy(),
+                    ]
+                )
+            else:
+                yerr = None
+            return y, yerr
+        # no fallback guessing when schema is frozen
+        raise KeyError(f"'{col}' has no 'center' subcolumn in binned_scalars.")
+    else:
+        if col in binned.columns:
+            return binned[col].to_numpy(), None
+        raise KeyError(f"Column '{col}' not found in binned_scalars.")
+
+
+def _index_to_numeric(idx: pd.Index) -> np.ndarray:
+    """
+    Convert a pandas Index into numeric values.
+
+    - If the index is an IntervalIndex, use midpoints.
+    - Otherwise, attempt direct float conversion.
+    - On failure, fallback to sequential integers.
+
+    Parameters
+    ----------
+    idx : pandas.Index
+        Input index to convert.
+
+    Returns
+    -------
+    values : numpy.ndarray
+        Numeric representation of the index.
+    """
+    if isinstance(idx, pd.IntervalIndex):
+        return ((idx.left.astype(float) + idx.right.astype(float)) / 2.0).to_numpy()
+    try:
+        return idx.to_numpy(dtype=float, copy=False)
+    except Exception:
+        return np.arange(len(idx), dtype=float)
+
+
+def plot_binned(
+    binned: pd.DataFrame,
+    x_col: Optional[str],
+    y_col: str,
+    *,
+    use_index_as_x: bool = False,
+    ax: Optional[plt.Axes] = None,
+    marker: str = "o",
+    linestyle: str = "-",
+    label: Optional[str] = None,
+    xscale: str = "linear",
+    yscale: str = "linear",
+    grid: bool = True,
+) -> plt.Axes:
+    """
+    Plot a single y-series versus x from binned data.
+
+    Parameters
+    ----------
+    binned : pandas.DataFrame
+        DataFrame of binned scalar values, with frozen schema.
+    x_col : str or None
+        Column to use for the x-axis. If None or `use_index_as_x=True`,
+        use the DataFrame index.
+    y_col : str
+        Column to plot as y-axis values. Must exist as 'center' in MultiIndex
+        schema or as a flat column.
+    use_index_as_x : bool, default=False
+        If True, use the DataFrame index for x-values regardless of `x_col`.
+    ax : matplotlib.axes.Axes, optional
+        Axes to plot into. If None, create a new figure and axes.
+    marker : str, default="o"
+        Marker style for the plot.
+    linestyle : str, default="-"
+        Line style for the plot.
+    label : str, optional
+        Label for the plotted series. If provided, a legend is added.
+    xscale : {"linear", "log"}, default="linear"
+        X-axis scale type.
+    yscale : {"linear", "log"}, default="linear"
+        Y-axis scale type.
+    grid : bool, default=True
+        Whether to draw a grid.
+
+    Returns
+    -------
+    ax : matplotlib.axes.Axes
+        The matplotlib Axes containing the plot.
+
+    Raises
+    ------
+    KeyError
+        If requested columns are not found in the schema.
+    """
+    # y (required)
+    y, yerr = _get_center_and_err(binned, y_col)
+
+    # x
+    if use_index_as_x or x_col is None:
+        x = _index_to_numeric(binned.index)
+        xerr = None
+        xlabel = binned.index.name or "bin"
+    else:
+        x, xerr = _get_center_and_err(binned, x_col)
+        xlabel = x_col
+
+    # mask + sort
+    mask = np.isfinite(x) & np.isfinite(y)
+    x, y = x[mask], y[mask]
+    if xerr is not None:
+        xerr = xerr[:, mask]
+    if yerr is not None:
+        yerr = yerr[:, mask]
+
+    order = np.argsort(x)
+    x, y = x[order], y[order]
+    if xerr is not None:
+        xerr = xerr[:, order]
+    if yerr is not None:
+        yerr = yerr[:, order]
+
+    if ax is None:
+        _, ax = plt.subplots()
+    ax.errorbar(
+        x, y, xerr=xerr, yerr=yerr, fmt=marker, linestyle=linestyle, label=label
+    )
+    ax.set_xlabel(xlabel)
+    ax.set_ylabel(y_col)
+    ax.set_xscale(xscale)
+    ax.set_yscale(yscale)
+    if grid:
+        ax.grid(True, alpha=0.3)
+    if label:
+        ax.legend()
+    return ax
+
+
+def plot_binned_multi(
+    binned: pd.DataFrame,
+    x_col: Optional[str],
+    y_cols: Sequence[str],
+    *,
+    use_index_as_x: bool = False,
+    labels: Optional[Sequence[str]] = None,
+    ax: Optional[plt.Axes] = None,
+    markers: Optional[Sequence[str]] = None,
+    linestyles: Optional[Sequence[str]] = None,
+    xscale: str = "linear",
+    yscale: str = "linear",
+    grid: bool = True,
+) -> plt.Axes:
+    """
+    Overlay multiple y-series versus the same x-axis.
+
+    Parameters
+    ----------
+    binned : pandas.DataFrame
+        DataFrame of binned scalar values, with frozen schema.
+    x_col : str or None
+        Column to use for the x-axis. If None or `use_index_as_x=True`,
+        use the DataFrame index.
+    y_cols : sequence of str
+        List of column names to plot as y-axis values.
+    use_index_as_x : bool, default=False
+        If True, use the DataFrame index for x-values regardless of `x_col`.
+    labels : sequence of str, optional
+        Labels for the plotted series. Defaults to `y_cols`.
+    ax : matplotlib.axes.Axes, optional
+        Axes to plot into. If None, create a new figure and axes.
+    markers : sequence of str, optional
+        Markers for each series. Defaults to a set of common markers.
+    linestyles : sequence of str, optional
+        Linestyles for each series. Defaults to all solid lines.
+    xscale : {"linear", "log"}, default="linear"
+        X-axis scale type.
+    yscale : {"linear", "log"}, default="linear"
+        Y-axis scale type.
+    grid : bool, default=True
+        Whether to draw a grid.
+
+    Returns
+    -------
+    ax : matplotlib.axes.Axes
+        The matplotlib Axes containing the plot.
+
+    Raises
+    ------
+    KeyError
+        If requested columns are not found in the schema.
+    """
+    if ax is None:
+        _, ax = plt.subplots()
+    if labels is None:
+        labels = y_cols
+    if markers is None:
+        markers = ["o", "s", "D", "^", "v", "P", "X"]
+    if linestyles is None:
+        linestyles = ["-"] * len(y_cols)
+
+    # common x
+    if use_index_as_x or x_col is None:
+        x = _index_to_numeric(binned.index)
+        xerr = None
+        xlabel = binned.index.name or "bin"
+    else:
+        x, xerr = _get_center_and_err(binned, x_col)
+        xlabel = x_col
+    order = np.argsort(x)
+
+    for i, (yc, lab) in enumerate(zip(y_cols, labels)):
+        y, yerr = _get_center_and_err(binned, yc)
+        m = markers[i % len(markers)]
+        ls = linestyles[i % len(linestyles)]
+        ax.errorbar(
+            x[order],
+            y[order],
+            xerr=None if xerr is None else xerr[:, order],
+            yerr=None if yerr is None else yerr[:, order],
+            fmt=m,
+            linestyle=ls,
+            label=lab,
+        )
+
+    ax.set_xlabel(xlabel)
+    ax.set_ylabel(", ".join(y_cols) if len(y_cols) == 1 else "value")
+    ax.set_xscale(xscale)
+    ax.set_yscale(yscale)
+    if grid:
+        ax.grid(True, alpha=0.3)
+    ax.legend()
+    return ax