Better Pandas

6.12. Better Pandas#

This section cover tools to make your experience with Pandas a litte bit better.

6.12.1. tqdm: Add Progress Bar to Your Pandas Apply#

If you want to keep informed about the progress of a pandas apply operation, use tqdm.

import pandas as pd 
from tqdm import tqdm 
import time 

df = pd.DataFrame({'a': [1, 2, 3, 4, 5], 'b': [2, 3, 4, 5, 6]})

tqdm.pandas()
def func(row):
    time.sleep(1)
    return row + 1

df['a'].progress_apply(func)

100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 5/5 [00:05<00:00,  1.00s/it]

  2
  3
  4
  5
  6
Name: a, dtype: int64

Link to tqdm.

6.12.2. pandarallel: A Simple Tool to Parallelize Pandas Operations#

If you want to parallelize your Pandas operations on all available CPUs by adding only one line of code, try pandarallel.

from pandarallel import pandarallel
import pandas as pd
from numpy.random import randint

df = pd.DataFrame(
    {
        "a": randint(0, 100, size=10000),
        "b": randint(0, 100, size=10000),
        "c": randint(0, 100, size=10000),
    }
)

pandarallel.initialize(progress_bar=True)
df.parallel_apply(lambda x: x**2)

INFO: Pandarallel will run on 8 workers.
INFO: Pandarallel will use standard multiprocessing data transfer (pipe) to transfer data between the main process and workers.

	a	b	c
0	3025	324	441
1	1	6561	5329
2	2025	4900	1024
3	25	5776	25
4	16	8100	3364
...	...	...	...
9995	49	676	4761
9996	3721	6889	4
9997	4225	9025	1156
9998	361	9	529
9999	5041	25	81

10000 rows × 3 columns

Link to pandarallel.

6.12.3. PandasAI: Gain Insights From Your pandas DataFrame With AI#

If you want to quickly gain insights from your pandas DataFrame with AI, use PandasAI. PandasAI serves as:

A tool to analyze your DataFrame
Not a tool to process your DataFrame

import pandas as pd  

df = pd.read_csv("https://raw.githubusercontent.com/mwaskom/seaborn-data/master/flights.csv")
df.head(10)

	year	month	passengers
0	1949	January	112
1	1949	February	118
2	1949	March	132
3	1949	April	129
4	1949	May	121
5	1949	June	135
6	1949	July	148
7	1949	August	148
8	1949	September	136
9	1949	October	119

print(df.head(5).to_markdown())

|    | year                | month    |   passengers |
|---:|:--------------------|:---------|-------------:|
|  0 | 1949-01-01 00:00:00 | January  |          112 |
|  1 | 1949-01-01 00:00:00 | February |          118 |
|  2 | 1949-01-01 00:00:00 | March    |          132 |
|  3 | 1949-01-01 00:00:00 | April    |          129 |
|  4 | 1949-01-01 00:00:00 | May      |          121 |

from pandasai import PandasAI
from pandasai.llm.openai import OpenAI

# Instantiate a LLM
llm = OpenAI(api_token="YOUR_API_TOKEN")

# Use pandasai
pandas_ai = PandasAI(llm, conversational=False)
print(
    pandas_ai.run(
        df,
        prompt="Which month of the years has the highest number of passengers on average?",
    )
)

The month with the highest average number of passengers is: July

print(
    pandas_ai.run(
        df, prompt="Which are the five years with the highest passenger numbers?"
    )
)

year
  5714
  5140
  4572
  4421
  3939
Name: passengers, dtype: int64

print(pandas_ai.run(df, prompt="Within what range of years does the dataset span?"))

        year     month  passengers
1949-01-01   January         112
1949-01-01  February         118
1949-01-01     March         132
1949-01-01     April         129
1949-01-01       May         121
The dataset spans from 1949 to 1960.

Link to PandasAI.

6.12.4. fugue: Use pandas Functions on the Spark and Dask Engines.#

Wouldn’t it be nice if you can leverage Spark or Dask to parallelize data science workloads using pandas syntax? Fugue allows you to do exactly that.

Fugue provides the transform function allowing users to use pandas functions on the Spark and Dask engines.

import pandas as pd
from typing import Dict
from fugue import transform
from fugue_spark import SparkExecutionEngine

input_df = pd.DataFrame({"id": [0, 1, 2], "fruit": (["apple", "banana", "orange"])})
map_price = {"apple": 2, "banana": 1, "orange": 3}


def map_price_to_fruit(df: pd.DataFrame, mapping: dict) -> pd.DataFrame:
    df["price"] = df["fruit"].map(mapping)
    return df


df = transform(
    input_df,
    map_price_to_fruit,
    schema="*, price:int",
    params=dict(mapping=map_price),
    engine=SparkExecutionEngine,
)
df.show()

21/10/01 11:17:05 WARN Utils: Your hostname, khuyen-Precision-7740 resolves to a loopback address: 127.0.1.1; using 192.168.1.90 instead (on interface wlp111s0)
21/10/01 11:17:05 WARN Utils: Set SPARK_LOCAL_IP if you need to bind to another address
WARNING: An illegal reflective access operation has occurred
WARNING: Illegal reflective access by org.apache.spark.unsafe.Platform (file:/home/khuyen/book/venv/lib/python3.8/site-packages/pyspark/jars/spark-unsafe_2.12-3.1.2.jar) to constructor java.nio.DirectByteBuffer(long,int)
WARNING: Please consider reporting this to the maintainers of org.apache.spark.unsafe.Platform
WARNING: Use --illegal-access=warn to enable warnings of further illegal reflective access operations
WARNING: All illegal access operations will be denied in a future release
21/10/01 11:17:05 WARN NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
Using Spark's default log4j profile: org/apache/spark/log4j-defaults.properties
Setting default log level to "WARN".
To adjust logging level use sc.setLogLevel(newLevel). For SparkR, use setLogLevel(newLevel).
21/10/01 11:17:06 WARN Utils: Service 'SparkUI' could not bind on port 4040. Attempting port 4041.
[Stage 2:===============>                                          (3 + 8) / 11]

+---+------+-----+
| id| fruit|price|
+---+------+-----+
|  0| apple|    2|
|  1|banana|    1|
|  2|orange|    3|
+---+------+-----+

[Stage 2:==========================================>               (8 + 3) / 11]

                                                                                

Link to fugue.

6.12.5. Version Your Pandas DataFrame with Delta Lake#

Versioning your data is essential to undoing mistakes, preventing data loss, and ensuring reproducibility. Delta Lake makes it easy to version pandas DataFrames and review past changes for auditing and debugging purposes.

To version a pandas DataFrame with Delta Lake, start with writing out a pandas DataFrame to a Delta table.

import pandas as pd
import os
from deltalake.writer import write_deltalake

df = pd.DataFrame({"x": [1, 2, 3]})

# Write to a delta table 
table = "delta_lake"
os.makedirs(table, exist_ok=True)
write_deltalake(table, df)

Delta Lake stores the data in a Parquet file and maintains a transaction log that records the data operations, enabling time travel and versioning.

delta_lake:

 ├──  0-4719861e-1d3a-49f8-8870-225e4e46e3a0-0.parquet  
 └──  _delta_log/ 
 │  └────  00000000000000000000.json  

To load the Delta table as a pandas DataFrame, simply use the DeltaTable object:

from deltalake import DeltaTable

dt = DeltaTable(table)
dt.to_pandas()

	x
0	1
1	2
2	3

Let’s see what happens when we append another pandas DataFrame to the Delta table.

df2 = pd.DataFrame({"x": [8, 9, 10]})

write_deltalake(table, df2, mode="append")

# Create delta table
dt = DeltaTable(table)
print(dt.to_pandas().to_markdown())

|    |   x |
|---:|----:|
|  0 |   1 |
|  1 |   2 |
|  2 |   3 |
|  3 |   8 |
|  4 |   9 |
|  5 |  10 |

Our Delta table now has two versions. Version 0 contains the initial data and Version 1 includes the data that was appended.

To get the metadata of files that currently make up the current table such as creation time, size, and statistics, call the get_add_actions method.

print(dt.get_add_actions(flatten=True).to_pandas())

                                               path  size_bytes   
0  0-a6738752-efca-4577-8cbf-9c69b404f2ee-0.parquet        1654  \
1  1-7a6df896-715a-4d4a-b210-b12e3fe57bc6-0.parquet        1654   

        modification_time  data_change  num_records  null_count.x  min.x   
0 2023-06-21 14:07:14.479         True            3             0      1  \
1 2023-06-21 14:07:18.657         True            3             0      8   

   max.x  
0      3  
1     10  

To access prior versions, simply specify the version number when loading the Delta table:

# Read Version 0 of the dataset
dt0 = DeltaTable(table, version=0)
dt0.to_pandas()

	x
0	1
1	2
2	3

Link to delta-rs.

6.12.6. Optimize Query Speed with Data Partitioning#

Partitioning data allows queries to target specific segments rather than scanning the entire table, which speeds up data retrieval.

The following code uses Delta Lake to select partitions from a pandas DataFrame. Partitioned data loading is approximately 24.5 times faster than loading the complete dataset and then querying a particular subset

import pandas as pd
from deltalake.writer import write_deltalake
from deltalake import DeltaTable
from datetime import datetime
import numpy as np

# Create a DataFrame with hourly sales data for 2 million records
np.random.seed(0)  # For reproducibility

start_date = datetime(2023, 1, 1)
end_date = datetime(2023, 8, 31)
date_range = pd.date_range(start_date, end_date, freq='H')

data = {
    'datetime': date_range,
    'value': np.random.randint(100, 1000, len(date_range))
}

df = pd.DataFrame(data)
df['month'] = df['datetime'].dt.month
df['day'] = df['datetime'].dt.day
df['hour'] = df['datetime'].dt.hour

df[["month", "day", "hour", "value"]].head(5)

	month	day	hour	value
0	1	1	0	784
1	1	1	1	659
2	1	1	2	729
3	1	1	3	292
4	1	1	4	935

# Write to a Delta table
table_path = 'delta_lake'
write_deltalake(table_path, df)

%%timeit
# Load the data from the Delta table
DeltaTable(table_path).to_pandas().query("month == 1 & day == 1")

79.2 ms ± 2.62 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)

# Write to a Delta table
table_path = "delta_lake2"
write_deltalake(table_path, df, partition_by=["month", "day"])

%%timeit
# Load the data from the Delta table
DeltaTable(table_path).to_pandas([("month", "=", "1"), ("day", "=", "1")])

3.23 ms ± 181 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)

Link to delta-rs.

6.12.7. Overwrite Partitions of a pandas DataFrame#

If you need to modify a specific subset of your pandas DataFrame, such as yesterday’s data, it is not possible to overwrite only that partition. Instead, you have to load the entire DataFrame into memory as a workaround solution.

Delta Lake makes it easy to overwrite partitions of a pandas DataFrame.

First, write out a pandas DataFrame as a Delta table that is partitioned by the date column.

import pandas as pd
from deltalake.writer import write_deltalake
from deltalake import DeltaTable

table_path = "tmp/records" 
df = pd.DataFrame(
    {"a": [1, 2, 3], "date": ["04-21", "04-22", "04-22"]}
)
write_deltalake(
    table_path,
    df,
    partition_by=["date"],
)

The Delta table’s contents are partitioned by date, with each partition represented by a directory

 └──  _delta_log/ 
 │  └────  00000000000000000000.json  
 └──  date=04-21/ 
 │  └────  0-a6813d0c-157b-4ca6-8b3c-8d5afd51947c-0.parquet  
 └──  date=04-22/ 
 │  └────  0-a6813d0c-157b-4ca6-8b3c-8d5afd51947c-0.parquet  

View the Delta table as a pandas DataFrame:

DeltaTable(table_path).to_pandas()

	a	date
0	2	04-22
1	3	04-22
2	1	04-21

Next, create another DataFrame with two other records on 04-22. Overwrite the 04-22 partition with the new DataFrame and leave other partitions untouched.

df = pd.DataFrame(
    {"a": [7, 8], "date": ["04-22", "04-22"]}
)
write_deltalake(
    table_path,
    df,
    mode="overwrite",
    partition_filters=[("date", "=", "04-22")],
)

DeltaTable(table_path).to_pandas()

	a	date
0	1	04-21
1	7	04-22
2	8	04-22

Here is the updated contents of the Delta table:

 └──  _delta_log/ 
 │  └────  00000000000000000000.json
 │  └────  00000000000000000001.json    
 └──  date=04-21/ 
 │  └────  0-a6813d0c-157b-4ca6-8b3c-8d5afd51947c-0.parquet  
 └──  date=04-22/ 
 │  ├────  0-a6813d0c-157b-4ca6-8b3c-8d5afd51947c-0.parquet  
 │  └────  1-b5c9640f-f386-4754-b28f-90e361ab4320-0.parquet 

Since the data files are not physically removed from disk, you can time travel to the initial version of the data.

DeltaTable(table_path, version=0).to_pandas()

	a	date
0	2	04-22
1	3	04-22
2	1	04-21

Link to delta-rs.

6.12.8. Efficient Data Appending in Parquet Files: Delta Lake vs. Pandas#

Appending data to an existing Parquet file using pandas involves:

Loading the entire existing table into memory.
Merging the new data with the existing table.
Writing the merged data to the existing file.

This process can be time-consuming and memory-intensive.

import pandas as pd  

df1 = pd.DataFrame([
    (1, "John", 5000),
    (2, "Jane", 6000),
], columns=["employee_id", "employee_name", "salary"])

df2 = pd.DataFrame([
    (3, "Alex", 8000),
], columns=["employee_id", "employee_name", "salary"])

# Save to a parquet file
df1.to_parquet("data.parquet")

# Read the data
existing_data = pd.read_parquet("data.parquet")

# Concat two dataframes
df3 = pd.concat([df1, df2])

# Save to a file
df3.to_parquet("data.parquet")

Delta Lake offers a more efficient approach to handling this process. With Delta Lake, you can add, remove, or modify columns without the need to recreate the entire table.

Delta Lake is also built on top of the Parquet file format so it retains the efficiency and columnar storage benefits of Parquet.

from deltalake.writer import write_deltalake

table_path = "employees"

# Write to Delta Lake
write_deltalake(table_path, df1)

# Append to Delta Lake
write_deltalake(table_path, df2, mode="append")

Link to delta-rs.

6.12.9. Enforce Data Quality with Delta Lake Constraints#

Delta Lake provides a convenient way to enforce data quality by adding constraints to a table, ensuring that only valid and consistent data can be added.

In the provided code, attempting to add new data with a negative salary violates the constraint of a positive salary, and thus, the data is not added to the table.

import pandas as pd
from deltalake.writer import write_deltalake
from deltalake import DeltaTable

table_path = "delta_lake"

df1 = pd.DataFrame(
    [
        (1, "John", 5000),
        (2, "Jane", 6000),
    ],
    columns=["employee_id", "employee_name", "salary"],
)

write_deltalake(table_path, df1)

print(df1.to_markdown())

|    |   employee_id | employee_name   |   salary |
|---:|--------------:|:----------------|---------:|
|  0 |             1 | John            |     5000 |
|  1 |             2 | Jane            |     6000 |

table = DeltaTable(table_path)
table.alter.add_constraint({"salary_gt_0": "salary > 0"})

df2 = pd.DataFrame(
    [(3, "Alex", -200)],
    columns=["employee_id", "employee_name", "salary"],
)

write_deltalake(table, df2, mode="append", engine="rust")

DeltaProtocolError: Invariant violations: ["Check or Invariant (salary > 0) violated by value in row: [3, Alex, -200]"]

Link to delta-rs.

6.12.10. Efficient Data Updates and Scanning with Delta Lake#

Every time new data is appended to an existing Delta table, a new Parquet file is generated. This allows data to be ingested incrementally without having to rewrite the entire dataset.

As files accumulate, read operations may surge. The compact function merges small files into larger ones, enhancing scanning performance.

Combining incremental processing with the compact function enables efficient data updates and scans as your dataset expands.

import pandas as pd
from deltalake.writer import write_deltalake

table_path = 'delta_lake'
data_url = "https://gist.githubusercontent.com/khuyentran1401/458905fc5c630d7a1f7a510a04e5e0f9/raw/5b2d760011c9255a68eb08b83b3b8759ffa25d5c/data.csv"
dfs = pd.read_csv(data_url, chunksize=100)

for df in dfs:
    write_deltalake(table_path, df, mode="append")

from deltalake import DeltaTable

dt = DeltaTable(table_path)

%%timeit
df = dt.to_pandas()

30.6 ms ± 2.94 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)

dt.optimize.compact()

{'numFilesAdded': 1,
 'numFilesRemoved': 100,
 'filesAdded': {'min': 278115,
  'max': 278115,
  'avg': 278115.0,
  'totalFiles': 1,
  'totalSize': 278115},
 'filesRemoved': {'min': 5712,
  'max': 5717,
  'avg': 5715.8,
  'totalFiles': 100,
  'totalSize': 571580},
 'partitionsOptimized': 1,
 'numBatches': 100,
 'totalConsideredFiles': 100,
 'totalFilesSkipped': 0,
 'preserveInsertionOrder': True}

%%timeit
df = dt.to_pandas()

1.32 ms ± 49 µs per loop (mean ± std. dev. of 7 runs, 1,000 loops each)

Link to delta-rs.

6.12.11. Simplify Table Merge Operations with Delta Lake#

Merging two datasets and performing both insert and update operations can be a complex task.

Delta Lake makes it easy to perform multiple data manipulation operations during a merge operation.

The following code demonstrates merging two datasets using Delta Lake:

If a match is found, the last_talk column in people_table is updated with the corresponding value from new_df.
If the last_talk value in people_table is older than 30 days and the corresponding row is not present in the new_df table, the status column is updated to ‘rejected’.

# Target table
people_table.toDF().show()

+---+-------+----------+------------+
| id|company| last_talk|      status|
+---+-------+----------+------------+
|  0|      A|2023-05-07|interviewing|
|  1|      B|2023-05-01|interviewing|
|  2|      C|2023-03-01|    rejected|
+---+-------+----------+------------+

# Source table
new_df.show()

+---+-------+----------+
| id|company| last_talk|
+---+-------+----------+
|  0|      A|2023-05-07|
+---+-------+----------+

one_month_ago = "current_date() - INTERVAL '30' DAY"

people_table.alias("target").merge(
    new_df.alias("source"), "target.id = source.id"
).whenMatchedUpdate(
    set={"target.last_talk": "source.last_talk", "target.status": "'interviewing'"}
).whenNotMatchedBySourceUpdate(
    condition=f"target.last_talk <= {one_month_ago}",
    set={"target.status": "'rejected'"},
).execute()

people_table.toDF().show()

+---+-------+----------+------------+
| id|company| last_talk|      status|
+---+-------+----------+------------+
|  0|      A|2023-05-07|interviewing|
|  1|      B|2023-05-01|interviewing|
|  2|      C|2023-03-01|    rejected|
+---+-------+----------+------------+

Link to Delta Lake.

6.12.12. The Best Way to Append Mismatched Data to Parquet Tables#

Appending mismatched data to a Parquet table involves reading the existing data, concatenating it with the new data, and overwriting the existing Parquet file. This approach can be expensive and may lead to schema inconsistencies.

In the following code, the datatype of col3 is supposed to be int64 instead of float64.

import pandas as pd  

filepath = 'test.parquet'

# Write a dataframe to a parquet file
df1 = pd.DataFrame({'col1': [1, 2], 'col2': [3, 4]})
df1.to_parquet(filepath)

# Append a dataframe to a parquet file
df2 = pd.DataFrame({'col1': [2], 'col2': [7], 'col3': [0]})
concatenation = pd.concat([df1, df2]) # concatenate dataframes
concatenation.to_parquet(filepath) # overwrite original file

concat_df = pd.read_parquet(filepath)
print(concat_df, "\n")
print(concat_df.dtypes)

   col1  col2  col3
0     1     3   NaN
1     2     4   NaN
0     2     7   0.0 

col1      int64
col2      int64
col3    float64
dtype: object

With Delta Lake, you can effortlessly append DataFrames with extra columns while ensuring the preservation of your data’s schema.

# Create a spark Dataframe
data = [(1, 3), (2, 4)]

df1 = (
    spark.createDataFrame(data)
    .toDF("col1", "col2")
    .repartition(1)
)

# Write to a delta table
path = "tmp"
df1.write.format("delta").save(path)

# Create a new DataFrame
new_data = [(2, 7, 0)]
df2 = (
    spark.createDataFrame(new_data).toDF("col1", "col2", "col3").repartition(1)
)
df2.show()

+----+----+----+
|col1|col2|col3|
+----+----+----+
|   2|   7|   0|
+----+----+----+

# Append to the existing Delta table
df2.write.option("mergeSchema", "true").mode("append").format("delta").save(path)

# Read the Delta table
from delta.tables import DeltaTable

table = DeltaTable.forPath(spark, path)
concat_df = table.toDF().pandas_api()

print(concat_df, "\n")
print(concat_df.dtypes)

   col1  col2  col3
0     2     7   0.0
1     1     3   NaN
2     2     4   NaN 

col1    int64
col2    int64
col3    int64
dtype: object

Link to Delta Lake.

6.12.13. Polars: Blazing Fast DataFrame Library#

If you want data manipulation library that’s both fast and memory-efficient, try Polars. Polars provides a high-level API similar to Pandas but with better performance for large datasets.

The code below compares the performance of Polars and pandas.

import pandas as pd
import polars as pl
import numpy as np
import time

# Create two Pandas DataFrames with 1 million rows each
pandas_df1 = pd.DataFrame({
    'key': np.random.randint(0, 1000, size=1_000_000),
    'value1': np.random.rand(1_000_000)
})

pandas_df2 = pd.DataFrame({
    'key': np.random.randint(0, 1000, size=1_000_000),
    'value2': np.random.rand(1000000)
})

# Create two Polars DataFrames from the Pandas DataFrames
polars_df1 = pl.from_pandas(pandas_df1)
polars_df2 = pl.from_pandas(pandas_df2)

# Merge the two DataFrames on the 'key' column
start_time = time.time()
pandas_merged = pd.merge(pandas_df1, pandas_df2, on='key')
pandas_time = time.time() - start_time

start_time = time.time()
polars_merged = polars_df1.join(polars_df2, on='key')
polars_time = time.time() - start_time

print(f"Pandas time: {pandas_time:.6f} seconds")
print(f"Polars time: {polars_time:.6f} seconds")

Pandas time: 127.604390 seconds
Polars time: 41.079080 seconds

print(f"Polars is {pandas_time/polars_time:.2f} times faster than Pandas")

Polars is 3.11 times faster than Pandas

Link to polars

6.12.14. Polars: Speed Up Data Processing 12x with Lazy Execution#

Polars is a lightning-fast DataFrame library that utilizes all available cores on your machine.

Polars has two APIs: an eager API and a lazy API.

The eager execution is similar to Pandas, which executes code immediately.

In contrast, the lazy execution defers computations until the collect() method is called. This approach avoids unnecessary computations, making lazy execution potentially more efficient than eager execution.

The code following code shows filter operations on a DataFrame containing 10 million rows. Running polars with lazy execution is 12 times faster than using pandas.

Create a pandas DataFrame and filter the DataFrame.

import pandas as pd


df = pd.DataFrame(data)
df.head()

	Cat1	Cat2	Num1	Num2
0	c	a	40	7292
1	d	b	45	7849
2	a	a	93	6940
3	c	a	46	1265
4	c	a	98	2509

%timeit df[(df['Cat1'] == 'a') & (df['Cat2'] == 'b') & (df['Num1'] >= 70)]

706 ms ± 75.4 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)

Create a polars DataFrame and filter the DataFrame.

import polars as pl

pl_df = pl.DataFrame(data)

%timeit pl_df.lazy().filter((pl.col('Cat1') == 'a') & (pl.col('Cat2') == 'b') & (pl.col('Num1') >= 70)).collect()

58.1 ms ± 428 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)

Link to polars

6.12.15. Polars vs. Pandas for CSV Loading and Filtering#

The read_csv method in Pandas loads all rows of the dataset into the DataFrame before filtering to remove all unwanted rows.

On the other hand, the scan_csv method in Polars delays execution and optimizes the operation until the collect method is called. This approach accelerates code execution, particularly when handling large datasets.

In the code below, it is 25.5 times faster to use Polars instead of Pandas to read a subset of CSV file containing 57k rows.

import pandas as pd
import polars as pl 

%%timeit
df = pd.read_csv("airport-codes.csv")
df[(df["type"] == "heliport") & (df["continent"] == "EU")]

143 ms ± 8.3 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)

%%timeit
pl.scan_csv("airport-codes.csv").filter(
    (pl.col("type") == "heliport") & (pl.col("continent") == "EU")
).collect()

5.6 ms ± 594 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)

6.12.16. Simple and Expressive Data Transformation with Polars#

Compared to pandas, Polars provides a more expressive syntax for creating complex data transformation pipelines. Every expression in Polars produces a new expression, and these expressions can be piped together.

import pandas as pd

df = pd.DataFrame(
    {"A": [1, 2, 6], "B": ["a", "b", "c"], "C": [True, False, True]}
)
integer_columns = df.select_dtypes("int64")
other_columns = df[["B"]]
pd.concat([integer_columns, other_columns], axis=1)

	A	B
0	1	a
1	2	b
2	6	c

import polars as pl

pl_df = pl.DataFrame(
    {"A": [1, 2, 6], "B": ["a", "b", "c"], "C": [True, False, True]}
)
pl_df.select([pl.col(pl.Int64), "B"])

shape: (3, 2)

A	B
i64	str
1	"a"
2	"b"
6	"c"

6.12.17. Harness Polars and Delta Lake for Blazing Fast Performance#

Polars is a DataFrame library written in Rust that has blazing fast performance. Delta Lake has helpful features including ACID transactions, time travel, schema enforcement, and more. Combining these two tools makes the code exceptionally powerful and efficient for data processing and analysis.

In the code below, using Polars for grouping operations is 6.88 times faster than using Pandas. Moreover, the integration of Delta Lake enables seamless time travel across different versions of the Polars DataFrame.

import pandas as pd
import numpy as np

# Number of rows in the dataset
num_rows = 10_000_000

# Creating categorical columns
category_col1 = np.random.choice(['A', 'B', 'C'], size=num_rows)

# Creating numerical columns
numeric_col1 = np.random.randint(low=1, high=100, size=num_rows)

# Creating the DataFrame
data = {
    'Cat1': category_col1,
    'Num1': numeric_col1,
}

df = pd.DataFrame(data)

# Outputting the first few rows of the dataset
df.tail()

	Cat1	Num1
9999995	B	42
9999996	A	23
9999997	A	79
9999998	B	87
9999999	B	19

pandas_time = %timeit -o df.groupby('Cat1').sum()

277 ms ± 55.7 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)

Create two versions of the Delta table:

from deltalake.writer import write_deltalake

save_path = "tmp/bear_delta_lake"

write_deltalake(save_path, df)

df2 = pd.DataFrame({"Cat1": ["B", "C"], "Num1": [2, 3]})

write_deltalake(save_path, df2, mode="append")

Read the latest version of the Delta table with Polars:

import polars as pl 

pl_df = pl.read_delta(save_path)
print(pl_df.tail())

shape: (5, 2)
┌──────┬──────┐
│ Cat1 ┆ Num1 │
│ ---  ┆ ---  │
│ str  ┆ i64  │
╞══════╪══════╡
│ A    ┆ 43   │
│ A    ┆ 14   │
│ C    ┆ 3    │
│ B    ┆ 2    │
│ C    ┆ 3    │
└──────┴──────┘

polars_time = %timeit -o pl_df.groupby("Cat1").sum()

40.3 ms ± 3.22 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)

Compare the performance of grouping operations using Pandas and Polars:

round(pandas_time.average/polars_time.average, 3)

6.882

print(f"Grouping operations using Polars is {pandas_time.average/polars_time.average:.2f} times faster than using Pandas.")

Grouping operations using Polars is 6.88 times faster than using Pandas.

Time travel to the version 0 of Delta table

print(pl.read_delta(save_path, version=0).tail())

shape: (5, 2)
┌──────┬──────┐
│ Cat1 ┆ Num1 │
│ ---  ┆ ---  │
│ str  ┆ i64  │
╞══════╪══════╡
│ C    ┆ 95   │
│ B    ┆ 38   │
│ A    ┆ 43   │
│ A    ┆ 14   │
│ C    ┆ 3    │
└──────┴──────┘

Link to polars

Link to delta-rs.

6.12.18. Parallel Execution of Multiple Files with Polars#

If you have multiple files to process, Polars enables you to construct a query plan for each file beforehand. This allows for the efficient execution of multiple files concurrently, maximizing processing speed.

import glob

import polars as pl

# Construct a query plan for each file
queries = []
for file in glob.glob("test_data/*.csv"):
    q = pl.scan_csv(file).group_by("Cat").agg(pl.sum("Num"))
    queries.append(q)

# Execute files in parallel
dataframes = pl.collect_all(queries)
dataframes

[shape: (3, 2)
 ┌─────┬─────┐
 │ Cat ┆ Num │
 │ --- ┆ --- │
 │ str ┆ i64 │
 ╞═════╪═════╡
 │ A   ┆ 2   │
 │ C   ┆ 6   │
 │ B   ┆ 4   │
 └─────┴─────┘,
 shape: (3, 2)
 ┌─────┬─────┐
 │ Cat ┆ Num │
 │ --- ┆ --- │
 │ str ┆ i64 │
 ╞═════╪═════╡
 │ B   ┆ 5   │
 │ A   ┆ 1   │
 │ C   ┆ 1   │
 └─────┴─────┘,
 shape: (3, 2)
 ┌─────┬─────┐
 │ Cat ┆ Num │
 │ --- ┆ --- │
 │ str ┆ i64 │
 ╞═════╪═════╡
 │ C   ┆ 4   │
 │ A   ┆ 4   │
 │ B   ┆ 1   │
 └─────┴─────┘]

Link to polars

Better Pandas

Contents

6.12. Better Pandas#

6.12.1. tqdm: Add Progress Bar to Your Pandas Apply#

6.12.2. pandarallel: A Simple Tool to Parallelize Pandas Operations#

6.12.3. PandasAI: Gain Insights From Your pandas DataFrame With AI#

6.12.4. fugue: Use pandas Functions on the Spark and Dask Engines.#

6.12.5. Version Your Pandas DataFrame with Delta Lake#

6.12.6. Optimize Query Speed with Data Partitioning#

6.12.7. Overwrite Partitions of a pandas DataFrame#

6.12.8. Efficient Data Appending in Parquet Files: Delta Lake vs. Pandas#

6.12.9. Enforce Data Quality with Delta Lake Constraints#

6.12.10. Efficient Data Updates and Scanning with Delta Lake#

6.12.11. Simplify Table Merge Operations with Delta Lake#

6.12.12. The Best Way to Append Mismatched Data to Parquet Tables#

6.12.13. Polars: Blazing Fast DataFrame Library#

6.12.14. Polars: Speed Up Data Processing 12x with Lazy Execution#

6.12.15. Polars vs. Pandas for CSV Loading and Filtering#

6.12.16. Simple and Expressive Data Transformation with Polars#

6.12.17. Harness Polars and Delta Lake for Blazing Fast Performance#

6.12.18. Parallel Execution of Multiple Files with Polars#