Feat/polars with columns , async with_columns pandas

docs/reference/decorators/with_columns.rst

@@ -2,17 +2,35 @@
 with_columns
 =======================
 
-Pandas
---------------
+We support the `with_columns` operation that appends the results as new columns to the original dataframe for several libraries:
 
-We have a ``with_columns`` option to run operations on columns of a Pandas dataframe and append the results as new columns.
+Pandas
+-----------------------
 
 **Reference Documentation**
 
 .. autoclass:: hamilton.plugins.h_pandas.with_columns
    :special-members: __init__
 
 
+Polar (Eager)
+-----------------------
+
+**Reference Documentation**
+
+.. autoclass:: hamilton.plugins.h_polars.with_columns
+   :special-members: __init__
+
+
+Polars (Lazy)
+-----------------------
+
+**Reference Documentation**
+
+.. autoclass:: hamilton.plugins.h_polars_lazyframe.with_columns
+   :special-members: __init__
+
+
 PySpark
 --------------
 

examples/pandas/with_columns/README

@@ -2,6 +2,6 @@
 
 We show the ability to use the familiar `with_columns` from either `pyspark` or `polars` on a Pandas dataframe.
 
-To see the example look at the notebook.
+To see the example look at the [notebook](notebook.ipynb).
 
 ![image info](./dag.png)

examples/polars/with_columns/README

@@ -0,0 +1,8 @@
+# Using with_columns with Polars
+
+We show the ability to use the familiar `with_columns` from `polars`. Supported for both: `pl.DataFrame` and `pl.LazyFrame`.
+
+To see the example look at the [notebook](notebook.ipynb).
+
+![image info](./DAG_DataFrame.png)
+![image info](./DAG_lazy.png)

examples/polars/with_columns/my_functions.py

@@ -0,0 +1,51 @@
+import polars as pl
+
+from hamilton.function_modifiers import config
+
+"""
+Notes:
+  1. This file is used for all the [ray|dask|spark]/hello_world examples.
+  2. It therefore show cases how you can write something once and not only scale it, but port it
+     to different frameworks with ease!
+"""
+
+
+@config.when(case="millions")
+def avg_3wk_spend__millions(spend: pl.Series) -> pl.Series:
+    """Rolling 3 week average spend."""
+    return (
+        spend.to_frame("spend").select(pl.col("spend").rolling_mean(window_size=3) / 1e6)
+    ).to_series(0)
+
+
+@config.when(case="thousands")
+def avg_3wk_spend__thousands(spend: pl.Series) -> pl.Series:
+    """Rolling 3 week average spend."""
+    return (
+        spend.to_frame("spend").select(pl.col("spend").rolling_mean(window_size=3) / 1e3)
+    ).to_series(0)
+
+
+def spend_per_signup(spend: pl.Series, signups: pl.Series) -> pl.Series:
+    """The cost per signup in relation to spend."""
+    return spend / signups
+
+
+def spend_mean(spend: pl.Series) -> float:
+    """Shows function creating a scalar. In this case it computes the mean of the entire column."""
+    return spend.mean()
+
+
+def spend_zero_mean(spend: pl.Series, spend_mean: float) -> pl.Series:
+    """Shows function that takes a scalar. In this case to zero mean spend."""
+    return spend - spend_mean
+
+
+def spend_std_dev(spend: pl.Series) -> float:
+    """Function that computes the standard deviation of the spend column."""
+    return spend.std()
+
+
+def spend_zero_mean_unit_variance(spend_zero_mean: pl.Series, spend_std_dev: float) -> pl.Series:
+    """Function showing one way to make spend have zero mean and unit variance."""
+    return spend_zero_mean / spend_std_dev

examples/polars/with_columns/my_functions_lazy.py

@@ -0,0 +1,47 @@
+import polars as pl
+
+from hamilton.function_modifiers import config
+
+"""
+Notes:
+  1. This file is used for all the [ray|dask|spark]/hello_world examples.
+  2. It therefore show cases how you can write something once and not only scale it, but port it
+     to different frameworks with ease!
+"""
+
+
+@config.when(case="millions")
+def avg_3wk_spend__millions(spend: pl.Expr) -> pl.Expr:
+    """Rolling 3 week average spend."""
+    return spend.rolling_mean(window_size=3) / 1e6
+
+
+@config.when(case="thousands")
+def avg_3wk_spend__thousands(spend: pl.Expr) -> pl.Expr:
+    """Rolling 3 week average spend."""
+    return spend.rolling_mean(window_size=3) / 1e3
+
+
+def spend_per_signup(spend: pl.Expr, signups: pl.Expr) -> pl.Expr:
+    """The cost per signup in relation to spend."""
+    return spend / signups
+
+
+def spend_mean(spend: pl.Expr) -> float:
+    """Shows function creating a scalar. In this case it computes the mean of the entire column."""
+    return spend.mean()
+
+
+def spend_zero_mean(spend: pl.Expr, spend_mean: float) -> pl.Expr:
+    """Shows function that takes a scalar. In this case to zero mean spend."""
+    return spend - spend_mean
+
+
+def spend_std_dev(spend: pl.Expr) -> float:
+    """Function that computes the standard deviation of the spend column."""
+    return spend.std()
+
+
+def spend_zero_mean_unit_variance(spend_zero_mean: pl.Expr, spend_std_dev: float) -> pl.Expr:
+    """Function showing one way to make spend have zero mean and unit variance."""
+    return spend_zero_mean / spend_std_dev