From 84d14cf9d9dccc28dd116c7a78d380633f082aa1 Mon Sep 17 00:00:00 2001
From: Tomasini Luca <luca.tomasini@heig-vd.ch>
Date: Wed, 12 Feb 2025 11:51:55 +0100
Subject: [PATCH] Add numpy_function

---
 docs/modules.rst        |  3 ++-
 docs/numpy_function.rst |  7 ++++++
 pyproject.toml          |  1 +
 src/numpy_function.py   | 30 +++++++++++++++++++++++
 src/polars_function.py  | 53 +++++++++++++++++++++++++++++++++++++++++
 5 files changed, 93 insertions(+), 1 deletion(-)
 create mode 100644 docs/numpy_function.rst
 create mode 100644 src/numpy_function.py

diff --git a/docs/modules.rst b/docs/modules.rst
index 072fc23..fddb96f 100644
--- a/docs/modules.rst
+++ b/docs/modules.rst
@@ -6,6 +6,7 @@ Utility Functions documentation
 
    general_function
    polars_function
-   polars_shapely_function
    shapely_function
+   polars_shapely_function
+   numpy_function
    networkx_function
diff --git a/docs/numpy_function.rst b/docs/numpy_function.rst
new file mode 100644
index 0000000..f8f6010
--- /dev/null
+++ b/docs/numpy_function.rst
@@ -0,0 +1,7 @@
+numpy\_function
+=======================
+
+.. automodule:: numpy_function
+   :members:
+   :undoc-members:
+   :show-inheritance:
diff --git a/pyproject.toml b/pyproject.toml
index 89241df..8b52986 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -10,6 +10,7 @@ packages = [
     {include = "networkx_function.py", from="src"},
     {include = "shapely_function.py", from="src"},
     {include = "polars_shapely_function.py", from="src"},
+    {include = "numpy_function.py", from="src"},
 ]  
 
 
diff --git a/src/numpy_function.py b/src/numpy_function.py
new file mode 100644
index 0000000..a0f7dda
--- /dev/null
+++ b/src/numpy_function.py
@@ -0,0 +1,30 @@
+import numpy as np
+
+def relative_error_within_boundaries(x: np.array, low: np.array, high: np.array) -> np.array: # type: ignore
+    """
+    Calculate the relative error of values within specified boundaries.
+
+    Args:
+        x (np.array): The array of values.
+        low (np.array): The lower boundary array.
+        high (np.array): The upper boundary array.
+
+    Returns:
+        np.array: The array of relative errors.
+    """
+    return np.abs(error_within_boundaries(x, low, high))/x
+
+def error_within_boundaries(x: np.array, low: np.array, high: np.array) -> np.array: # type: ignore
+    """
+    Calculate the error of values within specified boundaries.
+
+    Args:
+        x (np.array): The array of values.
+        low (np.array): The lower boundary array.
+        high (np.array): The upper boundary array.
+
+    Returns:
+        np.array: The array of errors.
+    """
+    nearest_boundary = np.where(x < low, low, np.where(x > high, high, x))
+    return x - nearest_boundary
\ No newline at end of file
diff --git a/src/polars_function.py b/src/polars_function.py
index 2e1ce66..2cab817 100644
--- a/src/polars_function.py
+++ b/src/polars_function.py
@@ -360,3 +360,56 @@ def concat_list_of_list(col_list: pl.Expr) -> pl.Expr:
     return pl.concat_list(
         col_list.map_elements(lambda x: [x], return_dtype=pl.List(pl.List(pl.Float64)))
     )
+
+
+def linear_interpolation_for_bound(x_col: pl.Expr, y_col: pl.Expr) -> pl.Expr:
+    """
+    Perform linear interpolation for boundary values in a column.
+
+    Args:
+        x_col (pl.Expr): The x-axis column.
+        y_col (pl.Expr): The y-axis column to interpolate.
+
+    Returns:
+        pl.Expr: The interpolated y-axis column.
+    """
+    a_diff: pl.Expr = y_col.diff()/x_col.diff()
+    x_diff: pl.Expr = x_col.diff().backward_fill()
+    y_diff: pl.Expr = pl.coalesce(
+        pl.when(y_col.is_null().or_(y_col.is_nan()))
+        .then(a_diff.forward_fill()*x_diff)
+        .otherwise(pl.lit(0)).cum_sum(),
+        pl.when(y_col.is_null().or_(y_col.is_nan()))
+        .then(-a_diff.backward_fill()*x_diff)
+        .otherwise(pl.lit(0)).cum_sum(reverse=True)
+    )
+
+    return y_col.backward_fill().forward_fill() + y_diff
+
+def linear_interpolation_using_cols(
+    df: pl.DataFrame, x_col: str, y_col: Union[list[str], str]
+    ) -> pl.DataFrame:
+    """
+    Perform linear interpolation on specified columns of a DataFrame.
+
+    Args:
+        df (pl.DataFrame): The DataFrame containing the data.
+        x_col (str): The name of the x-axis column.
+        y_col (Union[list[str], str]): The name(s) of the y-axis column(s) to interpolate.
+
+    Returns:
+        pl.DataFrame: The DataFrame with interpolated y-axis columns.
+    """
+    df = df.sort(x_col)
+    x = df[x_col].to_numpy()
+    if isinstance(y_col, str):
+        y_col = [y_col]
+    for col in y_col:
+        y = df[col].to_numpy()
+        mask = ~np.isnan(y)
+        df = df.with_columns(
+            pl.Series(np.interp(x, x[mask], y[mask], left=np.nan, right=np.nan)).fill_nan(None).alias(col)
+        ).with_columns(
+            linear_interpolation_for_bound(x_col=c(x_col), y_col=c(col)).alias(col)
+        )
+    return df
\ No newline at end of file