Add memory.py and task_memory.py to analyze stationarity and long-memory effects, providing a more detailed analysis of the AR(p) model's ability to represent the data.

Author: Lenr4

src/lennart_epp/analysis/memory.py

@@ -0,0 +1,78 @@
+import pandas as pd
+
+from lennart_epp.analysis.fit_ar_model import _check_stationarity, _difference_series
+
+error_msg = "Column '{column}' not found in dataframe."
+
+
+def check_stat_diff_close(df: pd.DataFrame, column: str = "close_price") -> dict:
+    """Check if the first-differenced close_price is stationary.
+
+    Args:
+        df (pd.DataFrame): The dataframe containing the time series.
+        column (str, optional): Column to test for stationarity.
+
+    Returns:
+        dict: A dictionary containing the ADF test results.
+    """
+    if column not in df.columns:
+        raise ValueError(error_msg)
+
+    df_diff = _difference_series(df, column)
+
+    diff_column = f"diff_{column}"
+
+    is_stationary, p_value, test_statistic_adf = _check_stationarity(
+        df_diff, diff_column
+    )
+
+    return {
+        "ADF Test Statistic": test_statistic_adf,
+        "P-Value": p_value,
+        "Is Stationary": is_stationary,
+    }
+
+
+def write_stationarity_results(results: dict, file_path):
+    """Generate and save a LaTeX table for Augmented Dickey-Fuller (ADF) test results.
+
+    Args:
+        results (dict): A dictionary containing the ADF test results with keys:
+            - "ADF Test Statistic" (float or None): The test statistic value.
+            - "P-Value" (float): The p-value from the ADF test.
+            - "Is Stationary" (bool): Whether the series is stationary.
+        file_path (Path): The file path where the LaTeX table will be saved.
+
+    Returns:
+        None: The function writes the LaTeX table to the specified file.
+    """
+    file_path.parent.mkdir(parents=True, exist_ok=True)
+
+    test_stat = results["ADF Test Statistic"]
+    test_stat_str = f"{test_stat:.4f}" if test_stat is not None else "Not Available"
+
+    conclusion_text = (
+        "The differenced series is likely stationary."
+        if results["Is Stationary"]
+        else "The differenced series may not be stationary."
+    )
+
+    latex_content = f"""
+    \\begin{{table}}[H]
+        \\centering
+        \\caption{{Results of the Augmented Dickey-Fuller (ADF) Test}}
+        \\label{{tab:stationarity_test}}
+        \\begin{{tabular}}{{l c}}
+            \\toprule
+            \\textbf{{Test Statistic}} & \\textbf{{Value}} \\\\
+            \\midrule
+            ADF Test Statistic & {test_stat_str} \\\\
+            P-Value & {results["P-Value"]:.4f} \\\\
+            Conclusion & {conclusion_text} \\\\
+            \\bottomrule
+        \\end{{tabular}}
+    \\end{{table}}
+    """
+
+    with file_path.open("w", encoding="utf-8") as f:
+        f.write(latex_content.strip())

src/lennart_epp/analysis/task_memory.py

@@ -0,0 +1,29 @@
+import pandas as pd
+
+from lennart_epp.analysis.memory import (
+    check_stat_diff_close,
+    write_stationarity_results,
+)
+from lennart_epp.config import BLD
+
+
+def task_check_stat_diff_close(
+    data=BLD / "data" / "cleaned_apple_data.pkl",
+    produces=BLD / "memory" / "diff_close_stat_test.tex",
+):
+    """Task to check if the differenced 'close_price' is stationary, and saves as .tex.
+
+    Args:
+        data (Path): Path to the cleaned Apple stock data (Pickle file).
+        produces (Path): Path to the output LaTeX file.
+
+    Returns:
+        None: Saves results to a .tex file.
+    """
+    df = pd.read_pickle(data)
+
+    results = check_stat_diff_close(df, column="close_price")
+
+    write_stationarity_results(results, produces)
+
+    assert produces.exists(), f" Failed to produce {produces}"