From 4fb3f070ab0c4af5f606a26f8b9ae879d4cd8801 Mon Sep 17 00:00:00 2001
From: Jerry Wu <jerry@ycwu.space>
Date: Tue, 17 Oct 2023 13:04:05 +0800
Subject: [PATCH] docs(python): Minor tweak in code example in section Coming
 from Pandas (#11764)

---
 docs/user-guide/migration/pandas.md | 25 +++++++++++--------------
 1 file changed, 11 insertions(+), 14 deletions(-)

diff --git a/docs/user-guide/migration/pandas.md b/docs/user-guide/migration/pandas.md
index d6674ac43f06..a9a039a7b1d4 100644
--- a/docs/user-guide/migration/pandas.md
+++ b/docs/user-guide/migration/pandas.md
@@ -147,19 +147,20 @@ called `hundredXValue` where the `value` column is multiplied by 100.
 In `Pandas` this would be:
 
 ```python
-df["tenXValue"] = df["value"] * 10
-df["hundredXValue"] = df["value"] * 100
+df.assign(
+    tenXValue=lambda df_: df_.value * 10,
+    hundredXValue=lambda df_: df_.value * 100
+)
 ```
 
 These column assignments are executed sequentially.
 
-In `Polars` we add columns to `df` using the `.with_columns` method and name them with
-the `.alias` method:
+In `Polars` we add columns to `df` using the `.with_columns` method:
 
 ```python
 df.with_columns(
-    (pl.col("value") * 10).alias("tenXValue"),
-    (pl.col("value") * 100).alias("hundredXValue"),
+    tenXValue=pl.col("value") * 10,
+    hundredXValue=pl.col("value") * 100,
 )
 ```
 
@@ -174,7 +175,7 @@ the values in column `a` based on a condition. When the value in column `c` is e
 In `Pandas` this would be:
 
 ```python
-df.loc[df["c"] == 2, "a"] = df.loc[df["c"] == 2, "b"]
+df.assign(a=lambda df_: df_.a.where(df_.c != 2, df_.b))
 ```
 
 while in `Polars` this would be:
@@ -187,21 +188,17 @@ df.with_columns(
 )
 ```
 
-The `Polars` way is pure in that the original `DataFrame` is not modified. The `mask` is
-also not computed twice as in `Pandas` (you could prevent this in `Pandas`, but that
-would require setting a temporary variable).
-
-Additionally `Polars` can compute every branch of an `if -> then -> otherwise` in
+`Polars` can compute every branch of an `if -> then -> otherwise` in
 parallel. This is valuable, when the branches get more expensive to compute.
 
 #### Filtering
 
 We want to filter the dataframe `df` with housing data based on some criteria.
 
-In `Pandas` you filter the dataframe by passing Boolean expressions to the `loc` method:
+In `Pandas` you filter the dataframe by passing Boolean expressions to the `query` method:
 
 ```python
-df.loc[(df['sqft_living'] > 2500) & (df['price'] < 300000)]
+df.query('m2_living > 2500 and price < 300000')
 ```
 
 while in `Polars` you call the `filter` method: