scitex 2.10.3__py3-none-any.whl → 2.13.0__py3-none-any.whl

scitex/plt/utils/_csv_column_naming.py

@@ -1,5 +1,4 @@
 #!/usr/bin/env python3
-# -*- coding: utf-8 -*-
 # Timestamp: 2025-12-08
 # File: ./src/scitex/plt/utils/_csv_column_naming.py
 
@@ -10,68 +9,68 @@ This module ensures consistent column naming between:
 - CSV export (_export_as_csv)
 - JSON metadata (_collect_figure_metadata)
 - GUI editors (reading CSV data back)
+- figrecipe compatibility
 
-Column naming convention:
-    {domain-name-value}_{domain-name-value}_...
-
-    - Underscore (_) separates different domains
-    - Hyphen (-) within domain name and between name-value
-
-Format:
-    ax-row-{row}-col-{col}_trace-id-{trace_id}_variable-{variable}
+Column naming convention (figrecipe-compatible):
+    r{row}c{col}_{caller}-{id}_{var}
 
 Where:
-    - ax-row-{row}-col-{col}: axes position in grid
-    - trace-id-{id}: unique identifier for the trace, which can be:
-        * User-provided id kwarg (e.g., "sine", "my-data")
-        * Generated from label (e.g., "sin-x" from "sin(x)")
-        * Auto-generated index (e.g., "0", "1")
-    - variable-{var}: type of data variable (e.g., "x", "y", "bins", "heights")
+    - r{row}c{col}: axes position in grid (e.g., r0c0, r1c2)
+    - {caller}: plotting method name (e.g., plot, scatter, bar)
+    - {id}: user-provided id kwarg OR auto-generated per-method counter
+    - {var}: variable name (e.g., x, y, bins, heights)
 
 Examples:
-    ax-row-0-col-0_trace-id-sine_variable-x       (row 0, col 0, id "sine", x)
-    ax-row-0-col-0_trace-id-sine_variable-y       (row 0, col 0, id "sine", y)
-    ax-row-0-col-1_trace-id-0_variable-x          (row 0, col 1, auto id 0, x)
-    ax-row-1-col-0_trace-id-my-data_variable-bins (row 1, col 0, "my-data", bins)
+    r0c0_plot-0_x        (row 0, col 0, plot method, auto-id 0, x variable)
+    r0c0_plot-sine_y     (row 0, col 0, plot method, id "sine", y variable)
+    r0c1_scatter-0_x     (row 0, col 1, scatter method, auto-id 0, x)
+    r1c0_bar-sales_height (row 1, col 0, bar method, id "sales", height)
+
+Legacy format (still supported for parsing):
+    ax-row-{row}-col-{col}_trace-id-{id}_variable-{var}
 """
 
+import re
+
 __all__ = [
     "get_csv_column_name",
     "get_csv_column_prefix",
     "parse_csv_column_name",
-    "sanitize_trace_id",
+    "sanitize_id",
     "get_unique_trace_id",
 ]
 
 
-def sanitize_trace_id(trace_id: str) -> str:
-    """Sanitize trace ID for use in CSV column names.
+def sanitize_id(raw_id: str) -> str:
+    """Sanitize ID for use in CSV column names.
 
     Removes or replaces characters that could cause issues in column names.
     Uses hyphen (-) for word separation within values.
 
     Parameters
     ----------
-    trace_id : str
-        Raw trace identifier (label, id kwarg, or generated)
+    raw_id : str
+        Raw identifier (label, id kwarg, or generated)
 
     Returns
     -------
     str
-        Sanitized trace ID safe for CSV column names
+        Sanitized ID safe for CSV column names
 
     Examples
     --------
-    >>> sanitize_trace_id("sin(x)")
+    >>> sanitize_id("sin(x)")
     'sin-x'
-    >>> sanitize_trace_id("My Data")
+    >>> sanitize_id("My Data")
     'my-data'
+    >>> sanitize_id("hello_world")
+    'hello-world'
     """
-    if not trace_id:
-        return "unnamed"
+    if not raw_id:
+        return "0"
 
-    # Replace problematic characters with hyphen (word separator within values)
-    sanitized = str(trace_id).lower()
+    # Replace problematic characters with hyphen
+    sanitized = str(raw_id).lower()
     result = []
     for char in sanitized:
         if char.isalnum():
@@ -89,16 +88,16 @@ def sanitize_trace_id(trace_id: str) -> str:
     # Remove leading/trailing hyphens
     sanitized = sanitized.strip("-")
 
-    return sanitized if sanitized else "unnamed"
+    return sanitized if sanitized else "0"
+
+
+# Backward compatibility alias
+sanitize_trace_id = sanitize_id
 
 
 def get_unique_trace_id(trace_id: str, existing_ids: set) -> str:
     """Get unique trace ID, adding suffix if collision detected.
 
-    This function ensures trace IDs remain unique even when multiple traces
-    have IDs that sanitize to the same value (e.g., "A" and "a" both become "a").
-    When a collision is detected, suffixes are added: a -> a-1 -> a-2, etc.
-
     Parameters
     ----------
     trace_id : str
@@ -116,20 +115,10 @@ def get_unique_trace_id(trace_id: str, existing_ids: set) -> str:
     >>> ids = set()
     >>> get_unique_trace_id("A", ids)
     'a'
-    >>> ids
-    {'a'}
     >>> get_unique_trace_id("a", ids)  # collision!
     'a-1'
-    >>> ids
-    {'a', 'a-1'}
-    >>> get_unique_trace_id("A ", ids)  # another collision!
-    'a-2'
-    >>> ids
-    {'a', 'a-1', 'a-2'}
-    >>> get_unique_trace_id("B", ids)  # no collision
-    'b'
     """
-    base_id = sanitize_trace_id(trace_id)
+    base_id = sanitize_id(trace_id)
 
     if base_id not in existing_ids:
         existing_ids.add(base_id)
@@ -146,13 +135,15 @@ def get_unique_trace_id(trace_id: str, existing_ids: set) -> str:
 
 
 def get_csv_column_prefix(
-    ax_row: int = 0, ax_col: int = 0, trace_id: str = None, trace_index: int = None
+    ax_row: int = 0,
+    ax_col: int = 0,
+    caller: str = "plot",
+    trace_id: str = None,
+    trace_index: int = None,
 ) -> str:
     """Get CSV column prefix for a trace.
 
-    Format: ax-row-{row}-col-{col}_trace-id-{id}_variable-
-    - Underscore (_) separates domains
-    - Hyphen (-) within domain names and between name-value
+    Format: r{row}c{col}_{caller}-{id}_
 
     Parameters
     ----------
@@ -160,45 +151,152 @@ def get_csv_column_prefix(
         Row position of axes in grid (default: 0)
     ax_col : int
         Column position of axes in grid (default: 0)
+    caller : str
+        Plotting method name (default: "plot")
     trace_id : str, optional
-        Trace identifier (from label or id kwarg). If None, uses trace_index.
+        User-provided trace identifier. If None, uses trace_index.
     trace_index : int, optional
-        Index of trace when no trace_id is provided (default: 0)
+        Auto-generated index when no trace_id provided (default: 0)
 
     Returns
     -------
     str
-        Column prefix like "ax-row-0-col-0_trace-id-sine_variable-"
+        Column prefix like "r0c0_plot-sine_"
 
     Examples
     --------
-    >>> get_csv_column_prefix(trace_id="sine")
-    'ax-row-0-col-0_trace-id-sine_variable-'
-    >>> get_csv_column_prefix(ax_row=1, ax_col=2, trace_index=0)
-    'ax-row-1-col-2_trace-id-0_variable-'
+    >>> get_csv_column_prefix(caller="plot", trace_id="sine")
+    'r0c0_plot-sine_'
+    >>> get_csv_column_prefix(ax_row=1, ax_col=2, caller="scatter", trace_index=0)
+    'r1c2_scatter-0_'
     """
     if trace_id:
-        safe_id = sanitize_trace_id(trace_id)
+        safe_id = sanitize_id(trace_id)
     elif trace_index is not None:
         safe_id = str(trace_index)
     else:
         safe_id = "0"
 
-    return f"ax-row-{ax_row}-col-{ax_col}_trace-id-{safe_id}_variable-"
+    return f"r{ax_row}c{ax_col}_{caller}-{safe_id}_"
+
+
+def _extract_caller_and_id(trace_id: str) -> tuple:
+    """Extract caller (method) and id from a combined trace_id.
+
+    Handles various formats:
+    - "plot_0" -> ("plot", "0")
+    - "scatter_1" -> ("scatter", "1")
+    - "stx_line_0" -> ("stx_line", "0")
+    - "sine" -> ("plot", "sine")  # user-provided, default to "plot"
+    - "plot-sine" -> ("plot", "sine")
+
+    Parameters
+    ----------
+    trace_id : str
+        Combined trace identifier
+
+    Returns
+    -------
+    tuple
+        (caller, id)
+    """
+    if not trace_id:
+        return ("plot", "0")
+
+    # Known method prefixes (order matters - longer first)
+    known_methods = [
+        "stx_line",
+        "stx_scatter",
+        "stx_bar",
+        "stx_barh",
+        "stx_box",
+        "stx_violin",
+        "stx_heatmap",
+        "stx_image",
+        "stx_imshow",
+        "stx_contour",
+        "stx_raster",
+        "stx_conf_mat",
+        "stx_joyplot",
+        "stx_rectangle",
+        "stx_fillv",
+        "stx_kde",
+        "stx_ecdf",
+        "stx_mean_std",
+        "stx_mean_ci",
+        "stx_median_iqr",
+        "stx_shaded_line",
+        "stx_errorbar",
+        "stx_fill_between",
+        "sns_boxplot",
+        "sns_violinplot",
+        "sns_barplot",
+        "sns_histplot",
+        "sns_kdeplot",
+        "sns_scatterplot",
+        "sns_lineplot",
+        "sns_swarmplot",
+        "sns_stripplot",
+        "sns_heatmap",
+        "sns_jointplot",
+        "sns_pairplot",
+        "plot_scatter",
+        "plot_box",
+        "plot_imshow",
+        "plot_kde",
+        "plot",
+        "scatter",
+        "bar",
+        "barh",
+        "hist",
+        "boxplot",
+        "violinplot",
+        "errorbar",
+        "fill_between",
+        "contour",
+        "contourf",
+        "imshow",
+        "pcolormesh",
+        "quiver",
+        "streamplot",
+        "stem",
+        "step",
+        "pie",
+        "hexbin",
+        "matshow",
+        "eventplot",
+        "stackplot",
+        "fill",
+        "text",
+        "annotate",
+    ]
+
+    # Check if trace_id starts with a known method
+    for method in known_methods:
+        # Check with underscore separator (e.g., "plot_0", "stx_line_0")
+        if trace_id.startswith(f"{method}_"):
+            remainder = trace_id[len(method) + 1 :]
+            return (method, remainder if remainder else "0")
+        # Check with hyphen separator (e.g., "plot-sine")
+        if trace_id.startswith(f"{method}-"):
+            remainder = trace_id[len(method) + 1 :]
+            return (method, remainder if remainder else "0")
+
+    # No known method prefix - assume user-provided ID, default caller to "plot"
+    return ("plot", trace_id)
 
 
 def get_csv_column_name(
     variable: str,
     ax_row: int = 0,
     ax_col: int = 0,
+    caller: str = None,
     trace_id: str = None,
     trace_index: int = None,
 ) -> str:
     """Get full CSV column name for a data field.
 
-    Format: ax-row-{row}-col-{col}_trace-id-{id}_variable-{var}
-    - Underscore (_) separates domains
-    - Hyphen (-) within domain names and between name-value
+    Format: r{row}c{col}_{caller}-{id}_{var}
 
     Parameters
     ----------
@@ -208,25 +306,34 @@ def get_csv_column_name(
         Row position of axes in grid (default: 0)
     ax_col : int
         Column position of axes in grid (default: 0)
+    caller : str, optional
+        Plotting method name. If None, extracted from trace_id or defaults to "plot"
     trace_id : str, optional
-        Trace identifier (from label or id kwarg)
+        User-provided trace identifier. May contain method prefix (e.g., "plot_0")
     trace_index : int, optional
-        Index of trace when no trace_id is provided
+        Auto-generated index when no trace_id provided
 
     Returns
     -------
     str
-        Full column name like "ax-row-0-col-0_trace-id-sine_variable-x"
+        Full column name like "r0c0_plot-sine_x"
 
     Examples
     --------
-    >>> get_csv_column_name("x", trace_id="sin(x)")
-    'ax-row-0-col-0_trace-id-sin-x_variable-x'
-    >>> get_csv_column_name("y", ax_row=1, ax_col=2, trace_index=0)
-    'ax-row-1-col-2_trace-id-0_variable-y'
+    >>> get_csv_column_name("x", caller="plot", trace_id="sine")
+    'r0c0_plot-sine_x'
+    >>> get_csv_column_name("y", ax_row=1, ax_col=2, caller="scatter", trace_index=0)
+    'r1c2_scatter-0_y'
+    >>> get_csv_column_name("x", trace_id="plot_0")  # backward compatible
+    'r0c0_plot-0_x'
     """
-    prefix = get_csv_column_prefix(ax_row, ax_col, trace_id, trace_index)
-    # Variable names are simple (x, y, bins, etc.)
+    # If caller not provided, try to extract from trace_id
+    if caller is None and trace_id:
+        caller, trace_id = _extract_caller_and_id(trace_id)
+    elif caller is None:
+        caller = "plot"
+
+    prefix = get_csv_column_prefix(ax_row, ax_col, caller, trace_id, trace_index)
     safe_variable = variable.lower()
     return f"{prefix}{safe_variable}"
 
@@ -234,14 +341,14 @@ def get_csv_column_name(
 def parse_csv_column_name(column_name: str) -> dict:
     """Parse CSV column name to extract components.
 
-    Format: ax-row-{row}-col-{col}_trace-id-{id}_variable-{var}
-    - Underscore (_) separates domains
-    - Hyphen (-) within domain names and between name-value
+    Supports both new format and legacy format:
+    - New: r{row}c{col}_{caller}-{id}_{var}
+    - Legacy: ax-row-{row}-col-{col}_trace-id-{id}_variable-{var}
 
     Parameters
     ----------
     column_name : str
-        Full column name (e.g., "ax-row-0-col-0_trace-id-sine_variable-x")
+        Full column name
 
     Returns
     -------
@@ -249,61 +356,73 @@ def parse_csv_column_name(column_name: str) -> dict:
         Dictionary with keys:
         - ax_row: int
         - ax_col: int
+        - caller: str (method name, empty for legacy format)
         - trace_id: str
         - variable: str
         - valid: bool (True if parsing succeeded)
 
     Examples
     --------
-    >>> parse_csv_column_name("ax-row-0-col-0_trace-id-sine_variable-x")
-    {'ax_row': 0, 'ax_col': 0, 'trace_id': 'sine', 'variable': 'x', 'valid': True}
-    >>> parse_csv_column_name("ax-row-1-col-2_trace-id-my-data_variable-bins")
-    {'ax_row': 1, 'ax_col': 2, 'trace_id': 'my-data', 'variable': 'bins', 'valid': True}
+    >>> parse_csv_column_name("r0c0_plot-sine_x")
+    {'ax_row': 0, 'ax_col': 0, 'caller': 'plot', 'trace_id': 'sine', 'variable': 'x', 'valid': True}
+    >>> parse_csv_column_name("r1c2_scatter-0_y")
+    {'ax_row': 1, 'ax_col': 2, 'caller': 'scatter', 'trace_id': '0', 'variable': 'y', 'valid': True}
     """
     result = {
         "ax_row": 0,
         "ax_col": 0,
+        "caller": "",
         "trace_id": "",
         "variable": "",
         "valid": False,
     }
 
-    if not column_name or not column_name.startswith("ax-row-"):
+    if not column_name:
+        return result
+
+    # Try new format: r{row}c{col}_{caller}-{id}_{var}
+    new_pattern = re.compile(r"^r(\d+)c(\d+)_([a-z_]+)-([^_]+)_([a-z]+)$")
+    match = new_pattern.match(column_name)
+    if match:
+        result["ax_row"] = int(match.group(1))
+        result["ax_col"] = int(match.group(2))
+        result["caller"] = match.group(3)
+        result["trace_id"] = match.group(4)
+        result["variable"] = match.group(5)
+        result["valid"] = True
         return result
 
-    try:
-        # Split by underscore to get domain groups
-        parts = column_name.split("_")
-        # Expected: ["ax-row-0-col-0", "trace-id-sine", "variable-x"]
-
-        for part in parts:
-            if part.startswith("ax-row-"):
-                # Parse ax-row-{row}-col-{col}
-                # Remove "ax-row-" prefix and split by "-col-"
-                rest = part[7:]  # Remove "ax-row-"
-                if "-col-" in rest:
-                    row_str, col_str = rest.split("-col-")
-                    result["ax_row"] = int(row_str)
-                    result["ax_col"] = int(col_str)
-            elif part.startswith("trace-id-"):
-                # Extract trace id (everything after "trace-id-")
-                result["trace_id"] = part[9:]
-            elif part.startswith("variable-"):
-                # Extract variable (everything after "variable-")
-                result["variable"] = part[9:]
-
-        # Validate we got all required fields
-        if result["variable"]:
-            result["valid"] = True
-
-    except (ValueError, IndexError):
-        pass
+    # Try legacy format: ax-row-{row}-col-{col}_trace-id-{id}_variable-{var}
+    if column_name.startswith("ax-row-"):
+        try:
+            parts = column_name.split("_")
+            for part in parts:
+                if part.startswith("ax-row-"):
+                    rest = part[7:]
+                    if "-col-" in rest:
+                        row_str, col_str = rest.split("-col-")
+                        result["ax_row"] = int(row_str)
+                        result["ax_col"] = int(col_str)
+                elif part.startswith("trace-id-"):
+                    result["trace_id"] = part[9:]
+                elif part.startswith("variable-"):
+                    result["variable"] = part[9:]
+
+            if result["variable"]:
+                result["valid"] = True
+        except (ValueError, IndexError):
+            pass
 
     return result
 
 
 def get_trace_columns_from_df(
-    df, trace_id: str = None, trace_index: int = None, ax_row: int = 0, ax_col: int = 0
+    df,
+    caller: str = None,
+    trace_id: str = None,
+    trace_index: int = None,
+    ax_row: int = 0,
+    ax_col: int = 0,
 ) -> dict:
     """Find CSV columns for a specific trace in a DataFrame.
 
@@ -311,6 +430,8 @@ def get_trace_columns_from_df(
     ----------
     df : pandas.DataFrame
         DataFrame with CSV data
+    caller : str, optional
+        Plotting method name to filter by
     trace_id : str, optional
         Trace identifier to search for
     trace_index : int, optional
@@ -323,18 +444,29 @@ def get_trace_columns_from_df(
     Returns
     -------
     dict
-        Dictionary mapping variable names to column names, e.g.:
-        {'x': 'ax-row-0-col-0_trace-id-sine_variable-x',
-         'y': 'ax-row-0-col-0_trace-id-sine_variable-y'}
+        Dictionary mapping variable names to column names
     """
     result = {}
-    prefix = get_csv_column_prefix(ax_row, ax_col, trace_id, trace_index)
 
-    for col in df.columns:
-        if col.startswith(prefix):
-            # Extract variable from column name
-            variable = col[len(prefix):]
-            result[variable] = col
+    if caller:
+        prefix = get_csv_column_prefix(ax_row, ax_col, caller, trace_id, trace_index)
+        for col in df.columns:
+            if col.startswith(prefix):
+                variable = col[len(prefix) :]
+                result[variable] = col
+    else:
+        # Search all columns matching position and trace
+        for col in df.columns:
+            parsed = parse_csv_column_name(col)
+            if (
+                parsed["valid"]
+                and parsed["ax_row"] == ax_row
+                and parsed["ax_col"] == ax_col
+            ):
+                if trace_id and parsed["trace_id"] == sanitize_id(trace_id):
+                    result[parsed["variable"]] = col
+                elif trace_index is not None and parsed["trace_id"] == str(trace_index):
+                    result[parsed["variable"]] = col
 
     return result
 

scitex/repro/README_RandomStateManager.md

@@ -28,7 +28,7 @@ import torch
 import tensorflow as tf
 
 # Initialize once - fixes everything!
-rng_manager = stx.rng.RandomStateManager(seed=42)
+rng = stx.rng.RandomStateManager(seed=42)
 
 # Python random - automatically fixed
 python_val = random.random()  # Reproducible!
@@ -87,7 +87,7 @@ import sys
 import scitex as stx
 
 # Session.start returns RNG, fixing random seeds of the supported libraries automatically
-CONFIG, stdout, stderr, plt, CC, rng_manager = stx.session.start(
+CONFIG, stdout, stderr, plt, CC, rng = stx.session.start(
     sys=sys,
     seed=42
 )
@@ -208,4 +208,4 @@ When you create a RandomStateManager, it automatically:
 
 yusuke.watanabe@scitex.ai
 
-<!-- EOF -->
+<!-- EOF -->