简体中文(ZH-CN) English(EN) 繁体中文(ZH-TW) 日本語(JA) 한국어(KO) Melayu(MS) Français(FR) Deutsch(DE)
집 > 백엔드 개발 > 파이썬 튜토리얼 > 본문

의사 결정 트리 알고리즘을 사용하여 호그와트 학생을 정렬하는 Gemika의 마법 가이드(파트 #7)

王林
풀어 주다： 2024-08-07 19:09:31
원래의
1056명이 탐색했습니다.

7. 원-핫 인코딩 주문을 시전하세요 ?✨

데이터 마법의 깊이를 탐구하는 동안 공기는 양피지 냄새와 마법 에너지의 탁탁거리는 소리로 가득 차 있습니다. 오늘은 원-핫 인코딩이라는 마법 결합 기술을 살펴보겠습니다. 활기 넘치는 성격과 예측할 수 없는 성격을 지닌 장난꾸러기 픽시를 마법 지도의 일련의 정확한 좌표로 변환하는 것을 상상해 보세요. 독자 여러분, 이것이 원-핫 인코딩의 핵심입니다. ?‍♂️✨

맥고나걸 교수의 지팡이가 찻잔을 닥스훈트로 바꿀 수 있는 것처럼, 이 주문은 범주형 데이터(수치 계산을 따르기를 거부하는 성가신 레이블)를 우리 모델이 이해할 수 있는 형식으로 바꿀 수 있습니다. Hufflepuffs, Ravenclaws, Gryffindors 및 Slytherins의 혼란스러운 떼를 각각 특정 집을 나타내는 1과 0의 깔끔한 격자로 바꾸는 것으로 생각하십시오. ?✨

원-핫 인코딩을 사용하면 각 고유 카테고리에 대해 새 열을 생성하고 해당 열을 1과 0으로 채워 존재 여부를 나타냅니다. 이는 장난꾸러기 집요정들을 지정된 방에 분류하여 각자가 자신만의 공간을 갖도록 하는 것과 같습니다. 이 장을 마치면 노련한 Charms 마스터의 자신감을 갖고 이 주문을 시전하여 데이터를 엉킨 매듭에서 아름답게 정리된 태피스트리로 변환할 수 있게 될 것입니다. ?✨

7.1 범주형 데이터: 모자와 마법의 라벨 정렬

숫자가 춤추고 패턴이 드러나는 데이터 과학의 매혹적인 영역에서 우리는 범주형 데이터라고 알려진 기이한 종류의 정보를 접하게 됩니다. 숫자 대응과 달리 이러한 데이터 포인트는 수량을 나타내지 않고 오히려 별개의 카테고리나 그룹을 나타냅니다.

호그와트의 분류 모자, 즉 학생들을 올바른 집에 배정하는 현명하고 오래된 마법의 물건을 상상해 보십시오. Gryffindor, Hufflepuff, Ravenclaw 및 Slytherin과 같은 주택은 범주형 데이터의 예입니다. 그들은 집 자체와 마찬가지로 독특한 특성을 지닌 별개의 그룹을 나타냅니다. 마찬가지로, 학생이 선택하는 애완동물의 유형(충성스러운 올빼미, 가르랑거리는 고양이, 심술궂은 두꺼비)도 범주형 데이터 범주에 속합니다.

범주형 데이터는 사물에 마법의 라벨을 붙이는 것과 같아서 분별 및 분류하는 데 도움이 됩니다. 약초학을 전공하는 학생이 속성에 따라 다양한 식물을 꼼꼼하게 분류하는 것처럼, 우리는 범주형 데이터를 사용하여 데이터 세트 내의 다양한 요소를 분류하고 이해합니다. 이러한 마법의 라벨을 이해함으로써 우리는 숨겨진 패턴을 풀고 강력한 주문(분석)을 시전하여 데이터의 비밀을 밝혀낼 수 있습니다. 하우스 열 아래에 어떤 가치가 있는지에 대한 추가 지식을 찾아보고 마법사 여러분의 지팡이를 던져 보세요. ?✨

# Displaying the unique categories in the 'House' column
unique_houses = hogwarts_df['house'].unique()
print(f"Unique Houses: {unique_houses}")
로그인 후 복사
Unique Houses: ['Gryffindor' 'Slytherin' 'Ravenclaw' 'Hufflepuff' 'Durmstrang' 'Beauxbatons']
로그인 후 복사

우리의 마법 알고리즘(또는 모델)은 이 데이터를 해석하는 방법을 알아야 하기 때문에 이러한 범주를 이해하는 것이 중요합니다. 그러나 이러한 알고리즘은 숫자에 더 익숙하기 때문에 숫자가 아닌 데이터로 인해 어려움을 겪는 경우가 많습니다. 원-핫 인코딩의 마법이 발휘되는 곳입니다.

7.2 원-핫 인코딩을 사용하여 범주형 데이터 변환

집, 지팡이 유형, 좋아하는 과목 등 매혹적인 세부 사항으로 가득 찬 호그와트 학생 기록을 상상해 보세요. 이러한 특성은 독특한 에너지를 전달하는 마법의 인장과 같습니다. 그러나 우리의 뛰어난 데이터 모델은 놀라운 업적을 달성할 수 있지만 이러한 인장을 직접 해독할 수는 없습니다. 우리는 그것을 그들이 이해할 수 있는 언어, 즉 숫자로 변환해야 합니다.

각 범주형 변수의 숨겨진 본질을 드러내는 강력한 주문, 원-핫 인코딩의 주문을 입력하세요. 마치 숨겨진 방에 루모스 주문을 걸어 구석구석을 밝혀주는 것과 같습니다. 코딩 막대를 가볍게 치면 각 카테고리를 자체 독립형 열로 변환합니다. 예를 들어, 학생이 그리핀도르에 속한 경우 마법처럼 그리핀도르 열에 1이 나타나고 다른 기숙사 열은 어두운 상태로 유지됩니다.

이러한 변형은 각 실이 카테고리를 나타내는 마법의 태피스트리를 만드는 것과 유사합니다. 이러한 스레드를 함께 엮음으로써 우리는 데이터 모델로 분석할 준비가 된 학생들의 풍부하고 상세한 초상화를 만듭니다. 이는 마치 우리 모델에게 폴리주스 포션 술꾼의 눈으로 세상을 볼 수 있는 능력을 부여하고 각 학생의 독특한 관점을 경험할 수 있는 능력을 부여하는 것과 같습니다. 첫 번째 열이나 기능을 One-Hot Encoding으로 시도해 보겠습니다. 성별 열을 먼저 시도해 보겠습니다.

# Importing necessary libraries
import pandas as pd
from sklearn.preprocessing import OneHotEncoder
from IPython.display import display, HTML

# Assuming hogwarts_df is already defined and contains the 'gender' column

# Applying One-Hot Encoding to the 'gender' column
encoder = OneHotEncoder(sparse_output=False)  # Updated parameter name
encoded_data = encoder.fit_transform(hogwarts_df[['gender']])

# Converting the encoded data into a DataFrame and attaching it to the original dataset
encoded_df = pd.DataFrame(encoded_data, columns=encoder.get_feature_names_out(['gender']))
hogwarts_df = pd.concat([hogwarts_df, encoded_df], axis=1)

# Dropping the original 'gender' column as it's now encoded
hogwarts_df.drop('gender', axis=1, inplace=True)

# Displaying the transformed DataFrame in a scrollable pane
html = hogwarts_df.head(5).to_html() # Convert DataFrame to HTML
scrollable_html = f"""
<div style="height: 300px; overflow: auto;">
    {html}
</div>
"""
display(HTML(scrollable_html))
로그인 후 복사
name    age origin  specialty   house   blood_status    pet wand_type   patronus    quidditch_position  boggart favorite_class  house_points    gender_Female   gender_Male
0   Harry Potter    11  England Defense Against the Dark Arts   Gryffindor  Half-blood  Owl Holly   Stag    Seeker  Dementor    Defense Against the Dark Arts   150.0   0.0 1.0
1   Hermione Granger    11  England Transfiguration Gryffindor  Muggle-born Cat Vine    Otter   Seeker  Failure Arithmancy  200.0   1.0 0.0
2   Ron Weasley 11  England Chess   Gryffindor  Pure-blood  Rat Ash Jack Russell Terrier    Keeper  Spider  Charms  50.0    0.0 1.0
3   Draco Malfoy    11  England Potions Slytherin   Pure-blood  Owl Hawthorn    Non-corporeal   Seeker  Lord Voldemort  Potions 100.0   0.0 1.0
4   Luna Lovegood   11  Ireland Creatures   Ravenclaw   Half-blood  Owl Fir Hare    Seeker  Her mother  Creatures   120.0   1.0 0.0
로그인 후 복사

And if you scroll to the right, you might notice that the dataset now has additional two columns, the gender_Female and the gender_Male on top of the existing one, while dropping the original gender column that was there previously.

7.3 The Two Great Treasures of the Data Realm ?✨

In the grand tapestry of the wizarding world of data, there exist two primary categories of magical artifacts: Structured Data and Unstructured Data. These are the building blocks of our enchanting spells and powerful potions.

Categorical Data is akin to a well-organized Herbology garden, where every plant (data point) has its rightful place. It's like a neatly filled Hogwarts student record, with columns for names, houses, and wand types, all aligned in perfect order. Structured data is a wizard's delight, easily understood and manipulated with a flick of the wand (or a few lines of code). ?✨

On the other hand, Numerical Data is a sprawling Forbidden Forest, filled with magical creatures (data points) roaming freely. It's like a collection of owls' letters, each with its own unique style and format. This data can be as diverse as the stars in the night sky, ranging from social media posts to news articles, images, and even spoken words. While it holds immense potential, taming this wild magic requires special spells and a keen eye for patterns. ??

  1. Categorical Data (Qualitative Data):

    • Definition: Categorical data refers to information that can be sorted into distinct groups or categories based on qualitative characteristics, rather than numerical values.
    • Types:
      • Nominal Data: This type includes categories without any specific order (e.g., gender, hair color). It is often used for labeling variables without providing a numerical value.
      • Ordinal Data: This type has a defined order or ranking (e.g., customer satisfaction ratings, economic class ratings, movie ratings). The differences between the ranks may not be equal.
    • Examples: Gender, race, color, and types of products.
    • Analysis: Categorical data is typically analyzed using frequency counts, bar graphs, and pie charts. It does not support arithmetic operations like addition or averaging.

  2. Numerical Data (Quantitative Data):

    • Definition: Numerical data consists of values that can be measured and expressed numerically, allowing for mathematical operations.
    • Types:
      • Discrete Data: Countable values (e.g., number of students in a class).
      • Continuous Data: Measurable quantities that can take any value within a range (e.g., height, weight).
    • Analysis: Numerical data can be analyzed using various statistical methods, including mean, median, mode, and standard deviation.

7.4 Transforming Text into Numbers: The Magic of One-Hot Encoding ✨

Imagine a world where numbers and words could converse, where the language of magic flowed seamlessly from one to the other. This is the realm of one-hot encoding, a powerful spell that transforms the mysterious world of text into the concrete world of numbers.

Just as a skilled Herbologist categorizes plants by their properties, one-hot encoding sorts textual data into distinct categories. Consider the Sorting Hat, which assigns students to houses based on their unique qualities. Similarly, one-hot encoding creates separate columns for each category, with values of 0 or 1 indicating whether a data point belongs to that category or not.

For instance, if you have a column representing the houses of Hogwarts students (Gryffindor, Slytherin, Ravenclaw, and Hufflepuff), one-hot encoding would conjure four new columns, one for each house. A value of 1 in the Gryffindor column would signify a student belonging to that house, while other columns would be filled with 0s. This numerical representation allows our magical models to understand and process textual information with ease. With that being said, let's try to transform the remaining of the columns in our dataset, and see what we have to work on. ?

# Importing necessary libraries
import pandas as pd
from sklearn.preprocessing import OneHotEncoder
from IPython.display import display, HTML

# Assuming hogwarts_df is already defined and contains the necessary columns
columns_to_encode = [
    'origin', 'specialty', 'blood_status', 'pet', 'wand_type', 'patronus', 'quidditch_position', 'boggart', 'favorite_class'
]

# Creating an instance of OneHotEncoder
encoder = OneHotEncoder(sparse_output=False)

# List to hold encoded DataFrames
encoded_dfs = []

# Applying One-Hot Encoding to each column and storing the result
for column in columns_to_encode:
    encoded_data = encoder.fit_transform(hogwarts_df[[column]])
    encoded_df = pd.DataFrame(encoded_data, columns=encoder.get_feature_names_out([column]))
    encoded_dfs.append(encoded_df)

# Concatenating all encoded DataFrames into one
encoded_df_combined = pd.concat(encoded_dfs, axis=1)

# Concatenating the encoded DataFrame with the original DataFrame
hogwarts_df = pd.concat([hogwarts_df, encoded_df_combined], axis=1)

# Dropping the original columns that were encoded
hogwarts_df.drop(columns=columns_to_encode, inplace=True)

# Displaying the transformed DataFrame in a scrollable pane
html = hogwarts_df.head(5).to_html()  # Convert DataFrame to HTML
scrollable_html = f"""
<div style="height: 300px; overflow: auto;">
    {html}
</div>
"""
display(HTML(scrollable_html))
로그인 후 복사
name    age house   house_points    gender_Female   gender_Male origin_Bulgaria origin_England  origin_Europe   origin_France   origin_Indonesia    origin_Ireland  origin_Scotland origin_USA  origin_Wales    specialty_Auror specialty_Baking    specialty_Charms    specialty_Chess specialty_Creatures specialty_Dark Arts specialty_Defense Against the Dark Arts specialty_Dueling   specialty_Goat Charming specialty_Gossip    specialty_Herbology specialty_History of Magic  specialty_Household Charms  specialty_Legilimency   specialty_Magical Creatures specialty_Memory Charms specialty_Metamorphmagus    specialty_Muggle Artifacts  specialty_Obscurus  specialty_Potions   specialty_Quidditch specialty_Strength  specialty_Transfiguration   specialty_Transformation    blood_status_Half-blood blood_status_Muggle-born    blood_status_No-mag blood_status_Pure-blood pet_Cat pet_Demiguise   pet_Dog pet_Goat    pet_Owl pet_Phoenix pet_Rat pet_Snake   pet_Toad    wand_type_Alder wand_type_Ash   wand_type_Birch wand_type_Blackthorn    wand_type_Cedar wand_type_Cherry    wand_type_Chestnut  wand_type_Cypress   wand_type_Ebony wand_type_Elder wand_type_Elm   wand_type_Fir   wand_type_Hawthorn  wand_type_Hazel wand_type_Hemlock   wand_type_Holly wand_type_Hornbeam  wand_type_Maple wand_type_Oak   wand_type_Pine  wand_type_Rosewood  wand_type_Rowan wand_type_Sword wand_type_Teak  wand_type_Vine  wand_type_Walnut    wand_type_Willow    wand_type_Yew   patronus_Cat    patronus_Doe    patronus_Dog    patronus_Eagle  patronus_Hare   patronus_Horse  patronus_Jack Russell Terrier   patronus_Lion   patronus_Non-corporeal  patronus_Otter  patronus_Phoenix    patronus_Serpent    patronus_Stag   patronus_Swan   patronus_Wolf   quidditch_position_Azkaban  quidditch_position_Beater   quidditch_position_Chaser   quidditch_position_Keeper   quidditch_position_Seeker   boggart_Ariana's death  boggart_Dementor    boggart_Dueling boggart_Failure boggart_Full Moon   boggart_Her mother  boggart_Lily Potter boggart_Lord Voldemort  boggart_Severus Snape   boggart_Spider  boggart_Tom Riddle  favorite_class_Arithmancy   favorite_class_Baking   favorite_class_Charms   favorite_class_Creatures    favorite_class_Dark Arts    favorite_class_Defense Against the Dark Arts    favorite_class_Dueling  favorite_class_Goat Charming    favorite_class_Gossip   favorite_class_Herbology    favorite_class_Household Charms favorite_class_Legilimency  favorite_class_Memory Charms    favorite_class_Muggle Studies   favorite_class_Obscurus favorite_class_Potions  favorite_class_Quidditch    favorite_class_Strength favorite_class_Transfiguration  favorite_class_Transformation
0   Harry Potter    11  Gryffindor  150.0   0.0 1.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
1   Hermione Granger    11  Gryffindor  200.0   1.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 1.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
2   Ron Weasley 11  Gryffindor  50.0    0.0 1.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
3   Draco Malfoy    11  Slytherin   100.0   0.0 1.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0
4   Luna Lovegood   11  Ravenclaw   120.0   1.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
로그인 후 복사

7.5 Discretizing the Numerical Values: Uncovering the Numerical Data) ✨

In the magical realm of data science, where numbers hold secrets and patterns dance in the shadows, there exists a particularly enchanting spell: one-hot encoding. This spell is a transfiguration charm, capable of transforming seemingly ordinary text into a numerical language that our magical computers can understand. Imagine a bustling Diagon Alley, filled with shops selling wands, cauldrons, and robes of every color. Each shop has a unique name, a string of letters that defines its identity. Now, picture these shop names as magical creatures, wild and untamed. To harness their power for our spells, we must transform them into something more manageable – numbers.

One-hot encoding is the spell that accomplishes this feat. It takes each unique shop name and creates a separate magical dimension (column) for it. Within these dimensions, we cast a binary spell, assigning a value of 1 to the shop that exists in that dimension and 0 to all others. It's like creating a magical grid, where each shop has its own spotlight moment. With this transformation, our once chaotic collection of shop names becomes an orderly array of numbers, ready to be analyzed and explored.?✨

7.5.1 Converting Columns with Numerical Values

To convert a numerical column with values ranging from 100 to 200 into a more machine learning-friendly format using one-hot encoding, you typically need to discretize the numerical values into categorical bins first. Here’s how to do it step by step:

You can create bins (categories) for the numerical values. For example, you might define bins like this:

  • 100-120
  • 121-140
  • 141-160
  • 161-180
  • 181-200

7.5.2 Assign Categories

Next, assign each numerical value to its corresponding bin. For example:

Original Value Category
100 100-120
110 100-120
125 121-140
145 141-160
165 161-180
180 161-180
200 181-200

7.5.3 One-Hot Encode the Categories

Now, you can apply one-hot encoding to the categorical column. Each category will be represented as a binary vector:

Original Value 100-120 121-140 141-160 161-180 181-200
100 1 0 0 0 0
110 1 0 0 0 0
125 0 1 0 0 0
145 0 0 1 0 0
165 0 0 0 1 0
180 0 0 0 1 0
200 0 0 0 0 1

By discretizing the numerical values into categories and applying one-hot encoding, you transform the original numerical data into a format that machine learning algorithms can process effectively. This method captures the categorical nature of the data while preserving the information contained in the original numerical values.

import pandas as pd
from sklearn.preprocessing import OneHotEncoder
from IPython.display import display, HTML

# Step 1: Define bins and labels
bins = [100, 120, 140, 160, 180, 200]  # Define the bin edges
labels = ['hp_100_120', 'hp_121_140', 'hp_141_160', 'hp_161_180', 'hp_181_200']  # Define the bin labels

# Step 2: Create a new categorical column based on the bins
hogwarts_df['house_category'] = pd.cut(hogwarts_df['house_points'], bins=bins, labels=labels, right=True)

# Step 3: One-hot encode the categorical column
hogwarts_df_encoded = pd.get_dummies(hogwarts_df, columns=['house_category'], prefix='', prefix_sep='')

# Replace True with 1 and False with 0
hogwarts_df_encoded = hogwarts_df_encoded.replace({True: 1, False: 0})

# Drop the house_points column
hogwarts_df_encoded.drop('house_points', axis=1, inplace=True)

# Displaying the transformed DataFrame in a scrollable pane
html = hogwarts_df_encoded.head(5).to_html()  # Convert DataFrame to HTML & # Display first 5 rows in a scrollable pane
scrollable_html = f"""
<div style="height: 300px; overflow: auto;">
    {html}
</div>
"""
display(HTML(scrollable_html))
로그인 후 복사
name    age house   gender_Female   gender_Male origin_Bulgaria origin_England  origin_Europe   origin_France   origin_Indonesia    origin_Ireland  origin_Scotland origin_USA  origin_Wales    specialty_Auror specialty_Baking    specialty_Charms    specialty_Chess specialty_Creatures specialty_Dark Arts specialty_Defense Against the Dark Arts specialty_Dueling   specialty_Goat Charming specialty_Gossip    specialty_Herbology specialty_History of Magic  specialty_Household Charms  specialty_Legilimency   specialty_Magical Creatures specialty_Memory Charms specialty_Metamorphmagus    specialty_Muggle Artifacts  specialty_Obscurus  specialty_Potions   specialty_Quidditch specialty_Strength  specialty_Transfiguration   specialty_Transformation    blood_status_Half-blood blood_status_Muggle-born    blood_status_No-mag blood_status_Pure-blood pet_Cat pet_Demiguise   pet_Dog pet_Goat    pet_Owl pet_Phoenix pet_Rat pet_Snake   pet_Toad    wand_type_Alder wand_type_Ash   wand_type_Birch wand_type_Blackthorn    wand_type_Cedar wand_type_Cherry    wand_type_Chestnut  wand_type_Cypress   wand_type_Ebony wand_type_Elder wand_type_Elm   wand_type_Fir   wand_type_Hawthorn  wand_type_Hazel wand_type_Hemlock   wand_type_Holly wand_type_Hornbeam  wand_type_Maple wand_type_Oak   wand_type_Pine  wand_type_Rosewood  wand_type_Rowan wand_type_Sword wand_type_Teak  wand_type_Vine  wand_type_Walnut    wand_type_Willow    wand_type_Yew   patronus_Cat    patronus_Doe    patronus_Dog    patronus_Eagle  patronus_Hare   patronus_Horse  patronus_Jack Russell Terrier   patronus_Lion   patronus_Non-corporeal  patronus_Otter  patronus_Phoenix    patronus_Serpent    patronus_Stag   patronus_Swan   patronus_Wolf   quidditch_position_Azkaban  quidditch_position_Beater   quidditch_position_Chaser   quidditch_position_Keeper   quidditch_position_Seeker   boggart_Ariana's death  boggart_Dementor    boggart_Dueling boggart_Failure boggart_Full Moon   boggart_Her mother  boggart_Lily Potter boggart_Lord Voldemort  boggart_Severus Snape   boggart_Spider  boggart_Tom Riddle  favorite_class_Arithmancy   favorite_class_Baking   favorite_class_Charms   favorite_class_Creatures    favorite_class_Dark Arts    favorite_class_Defense Against the Dark Arts    favorite_class_Dueling  favorite_class_Goat Charming    favorite_class_Gossip   favorite_class_Herbology    favorite_class_Household Charms favorite_class_Legilimency  favorite_class_Memory Charms    favorite_class_Muggle Studies   favorite_class_Obscurus favorite_class_Potions  favorite_class_Quidditch    favorite_class_Strength favorite_class_Transfiguration  favorite_class_Transformation   hp_100_120  hp_121_140  hp_141_160  hp_161_180  hp_181_200
0   Harry Potter    11  Gryffindor  0.0 1.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0   0   1   0   0
1   Hermione Granger    11  Gryffindor  1.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 1.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0   0   0   0   1
2   Ron Weasley 11  Gryffindor  0.0 1.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0   0   0   0   0
3   Draco Malfoy    11  Slytherin   0.0 1.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0   0   0   0   0
4   Luna Lovegood   11  Ravenclaw   1.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1   0   0   0   0
로그인 후 복사

7.6 Transforming Numbers into Magical Categories: Discretizing Age ?‍♀️✨

Imagine a bustling Hogwarts Express, filled with students of all ages. From wide-eyed first-years to wise-cracking seventh-years, each student possesses a unique blend of magic and mischief. To better understand these young wizards and witches, we can transform their ages from precise numbers into magical categories.

This process, known as discretization, is like sorting mischievous house-elves into their designated chores. Instead of treating age as a continuous spectrum, we group students into specific age ranges or bins. Picture these bins as enchanted compartments on the Hogwarts Express, each holding students of similar age.

By discretizing age, we unlock a new dimension of magical insight. We can explore how different age groups perform in classes, participate in Quidditch, or even their preferred wand type. It's like casting a revealing charm on our data, uncovering hidden patterns and trends that would otherwise remain obscured. So, let's transform those numbers into magical categories and discover the enchanting secrets hidden within the age of our Hogwarts students! ?✨

import pandas as pd
from IPython.display import display, HTML

# Step 1: Define bins and labels for the age column
bins = [10, 12, 14, 16, 18]  # Define the bin edges
labels = ['age_11', 'age_12', 'age_13', 'age_14']  # Define the bin labels

# Step 2: Create a new categorical column based on the bins
hogwarts_df_encoded['age_category'] = pd.cut(hogwarts_df_encoded['age'], bins=bins, labels=labels, right=True)

# Step 3: One-hot encode the categorical column
hogwarts_df_encoded_age = pd.get_dummies(hogwarts_df_encoded, columns=['age_category'], prefix='', prefix_sep='')

# Replace True with 1 and False with 0 (not necessary here since get_dummies already returns integers)
hogwarts_df_encoded_age = hogwarts_df_encoded_age.replace({True: 1, False: 0})

# Drop the age column
hogwarts_df_encoded_age.drop('age', axis=1, inplace=True)

# Displaying the transformed DataFrame in a scrollable pane
html = hogwarts_df_encoded_age.head(5).to_html()  # Convert DataFrame to HTML & Display first 5 rows in a scrollable pane
scrollable_html = f"""
<div style="height: 300px; overflow: auto;">
    {html}
</div>
"""
display(HTML(scrollable_html))
로그인 후 복사

7.7 Order in the Chaos: The Magic of Naming Conventions ? ✨

Imagine a bustling Hogwarts library, books overflowing from shelves, each with its own peculiar title. Without a proper cataloguing system, finding a specific spellbook would be like searching for a needle in a haystack, wouldn't it? In the realm of data, our columns are like these books. They hold valuable information, but without a clear and consistent naming system, they can quickly become a chaotic mess. This is where the magic of naming conventions comes into play.

Just as a librarian organizes books by subject, author, and title, we must impose order upon our columns. A well-crafted naming convention is like a powerful sorting spell, grouping similar columns together and making them easily identifiable. For example, using prefixes like 'student_', 'subject_', or 'score_' can instantly clarify the column's purpose.

By adopting a clear and consistent naming convention, you'll transform your data from a chaotic jumble into a well-organized magical library. This not only improves readability but also streamlines data manipulation and analysis. So, next time you're wrangling with data, remember the importance of a strong naming convention. It's the first step towards unlocking the hidden treasures within your dataset! ✨

import pandas as pd
from IPython.display import display, HTML

# Assuming hogwarts_df_encoded is already defined and contains the necessary columns

# Manipulate column names
hogwarts_df_encoded_age.columns = hogwarts_df_encoded_age.columns.str.lower().str.replace(' ', '_').str.replace('-', '-').str.replace("'", "")

# Display the transformed DataFrame in a scrollable pane
html = hogwarts_df_encoded_age.head(5).to_html()  # Convert first 5 rows to HTML
scrollable_html = f"""
<div style="height: 300px; overflow: auto;">
    {html}
</div>
"""
display(HTML(scrollable_html))  # Display first 5 rows in a scrollable pane
로그인 후 복사
name    house   gender_female   gender_male origin_bulgaria origin_england  origin_europe   origin_france   origin_indonesia    origin_ireland  origin_scotland origin_usa  origin_wales    specialty_auror specialty_baking    specialty_charms    specialty_chess specialty_creatures specialty_dark_arts specialty_defense_against_the_dark_arts specialty_dueling   specialty_goat_charming specialty_gossip    specialty_herbology specialty_history_of_magic  specialty_household_charms  specialty_legilimency   specialty_magical_creatures specialty_memory_charms specialty_metamorphmagus    specialty_muggle_artifacts  specialty_obscurus  specialty_potions   specialty_quidditch specialty_strength  specialty_transfiguration   specialty_transformation    blood_status_half-blood blood_status_muggle-born    blood_status_no-mag blood_status_pure-blood pet_cat pet_demiguise   pet_dog pet_goat    pet_owl pet_phoenix pet_rat pet_snake   pet_toad    wand_type_alder wand_type_ash   wand_type_birch wand_type_blackthorn    wand_type_cedar wand_type_cherry    wand_type_chestnut  wand_type_cypress   wand_type_ebony wand_type_elder wand_type_elm   wand_type_fir   wand_type_hawthorn  wand_type_hazel wand_type_hemlock   wand_type_holly wand_type_hornbeam  wand_type_maple wand_type_oak   wand_type_pine  wand_type_rosewood  wand_type_rowan wand_type_sword wand_type_teak  wand_type_vine  wand_type_walnut    wand_type_willow    wand_type_yew   patronus_cat    patronus_doe    patronus_dog    patronus_eagle  patronus_hare   patronus_horse  patronus_jack_russell_terrier   patronus_lion   patronus_non-corporeal  patronus_otter  patronus_phoenix    patronus_serpent    patronus_stag   patronus_swan   patronus_wolf   quidditch_position_azkaban  quidditch_position_beater   quidditch_position_chaser   quidditch_position_keeper   quidditch_position_seeker   boggart_arianas_death   boggart_dementor    boggart_dueling boggart_failure boggart_full_moon   boggart_her_mother  boggart_lily_potter boggart_lord_voldemort  boggart_severus_snape   boggart_spider  boggart_tom_riddle  favorite_class_arithmancy   favorite_class_baking   favorite_class_charms   favorite_class_creatures    favorite_class_dark_arts    favorite_class_defense_against_the_dark_arts    favorite_class_dueling  favorite_class_goat_charming    favorite_class_gossip   favorite_class_herbology    favorite_class_household_charms favorite_class_legilimency  favorite_class_memory_charms    favorite_class_muggle_studies   favorite_class_obscurus favorite_class_potions  favorite_class_quidditch    favorite_class_strength favorite_class_transfiguration  favorite_class_transformation   hp_100_120  hp_121_140  hp_141_160  hp_161_180  hp_181_200  age_11  age_12  age_13  age_14
0   Harry Potter    Gryffindor  0.0 1.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0   0   1   0   0   1   0   0   0
1   Hermione Granger    Gryffindor  1.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 1.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0   0   0   0   1   1   0   0   0
2   Ron Weasley Gryffindor  0.0 1.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0   0   0   0   0   1   0   0   0
3   Draco Malfoy    Slytherin   0.0 1.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0   0   0   0   0   1   0   0   0
4   Luna Lovegood   Ravenclaw   1.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1   0   0   0   0   1   0   0   0
로그인 후 복사

And once we've satisfied with the results, let's go ahead and save the current dataset to make the next enchanting journey more easier to navigate.

hogwarts_df_encoded_age.to_csv('data/hogwarts-students-03.csv', index=False)
로그인 후 복사

7.8 Mapping the Magical Connections: Uncovering Hidden Relationships

Imagine a vast enchanted forest, teeming with magical creatures and extraordinary plants. Each creature and plant possesses unique qualities, and some may share hidden connections. To unveil these mysterious bonds, we must cast a spell of correlation. A correlation matrix is like a magical map, guiding us through this enchanted forest. Each point on the map represents a different creature or plant, and the lines connecting them reveal the strength of their relationship. A thick, vibrant line signifies a strong connection, while a thin, faint line indicates a weaker bond.

As we explore this magical map, we search for patterns and trends. Are certain creatures often found near specific plants? Do particular plants thrive in the company of others? By deciphering these connections, we can uncover hidden knowledge about the forest and its inhabitants. Just as a skilled Herbologist studies the interactions between plants, we, as data wizards, uncover the hidden relationships between variables. With the correlation matrix as our guide, we embark on a thrilling adventure to explore the magical tapestry of our data! ?️✨

# Importing necessary libraries
import pandas as pd
from IPython.display import display, HTML

# Assuming hogwarts_df is already defined and contains the necessary columns

# Selecting only numerical columns
numerical_df = hogwarts_df_encoded_age.select_dtypes(include=['number'])

# Calculating the correlation matrix
correlation_matrix = numerical_df.corr()

# Displaying the correlation matrix in a scrollable pane
correlation_html = correlation_matrix.to_html()  # Convert correlation matrix to HTML
scrollable_correlation_html = f"""
<div style="height: 300px; overflow: auto;">
    {correlation_html}
</div>
"""
display(HTML(scrollable_correlation_html))  # Display correlation matrix in a scrollable pane
로그인 후 복사

The Gemika

And to make our analysis easier, let's just go ahead and save the correlation matrix into a tabular format.

correlation_matrix.to_csv('data/correlation-matrix.csv')
로그인 후 복사

7.9 Potion Ingredients: Verifying Our Data Types ?✨

Just as a skilled potioneer carefully examines their ingredients before brewing a powerful concoction, we data scientists must meticulously inspect our data types. These data types are like the magical properties of our ingredients, determining how they will react in our spells (algorithms).

Imagine our dataset as a cauldron brimming with magical elements. Each element, be it a student's age, house, or wand type, has a specific form or essence - its data type. These types can be as diverse as the magical creatures of the Forbidden Forest: numbers (like the count of house points), text (like a student's name), dates (like the founding year of Hogwarts), and more.

Mismatched data types can lead to disastrous results, like a potion exploding or a spell backfiring. That's why we must cast a discerning eye over our data, ensuring each element is of the correct type. It's like checking if a Mandrake root is truly a root and not a disguised Goblin! Only then can we confidently proceed with our magical data transformations. ?‍♀️✨

# Importing necessary libraries
import pandas as pd
from IPython.display import display, HTML

# Assuming hogwarts_df is already defined and contains the necessary columns

# Setting display options to show all columns and prevent truncation
pd.set_option('display.max_columns', None)  # Show all columns
pd.set_option('display.expand_frame_repr', False)  # Prevent truncation in output

# Checking the data types of each column
data_types_df = hogwarts_df_encoded_age.dtypes.to_frame(name='Data Type')  # Convert data types to a DataFrame

# Displaying the data types in a scrollable pane
data_types_html = data_types_df.to_html()  # Convert DataFrame to HTML
scrollable_data_types_html = f"""
<div style="height: 150px; overflow: auto;">
    {data_types_html}
</div>
"""
display(HTML(scrollable_data_types_html))  # Display data types in a scrollable pane
로그인 후 복사
    Data Type
name    object
house   object
gender_female   float64
gender_male float64
origin_bulgaria float64
origin_england  float64
origin_europe   float64
origin_france   float64
origin_indonesia    float64
origin_ireland  float64
origin_scotland float64
origin_usa  float64
origin_wales    float64
specialty_auror float64
specialty_baking    float64
specialty_charms    float64
specialty_chess float64
specialty_creatures float64
specialty_dark_arts float64
specialty_defense_against_the_dark_arts float64
specialty_dueling   float64
specialty_goat_charming float64
specialty_gossip    float64
specialty_herbology float64
specialty_history_of_magic  float64
specialty_household_charms  float64
specialty_legilimency   float64
specialty_magical_creatures float64
로그인 후 복사

7.10 Gemika's Pop-Up Quiz: Spotting the Trends

And now, dear apprentices, Gemika Haziq Nugroho has prepared a quiz to test your understanding of this powerful encoding spell. Are you ready to decode the mysteries of One-Hot Encoding?

  • What is categorical data, and why is it important in data analysis?
  • How does One-Hot Encoding transform categorical data for machine learning models?
  • Why do we drop the original categorical column after applying One-Hot Encoding?

Answer these questions to prove your mastery over the art of data transformation. As we continue our journey through the magical world of data science, remember that each spell we learn brings us closer to unveiling the secrets of our enchanted dataset. ??‍♂️✨

With these newfound skills, you're well-equipped to handle categorical data in any dataset. The path to becoming a master data wizard is full of wonder and discovery, and with each step, we draw closer to the heart of the magical data that surrounds us. Let us press on, eager to learn and ready to explore! ??

위 내용은 의사 결정 트리 알고리즘을 사용하여 호그와트 학생을 정렬하는 Gemika의 마법 가이드(파트 #7)의 상세 내용입니다. 자세한 내용은 PHP 중국어 웹사이트의 기타 관련 기사를 참조하세요!

원천：dev.to
이전 기사：Lyzr.ai로 텍스트 변환: 단계별 가이드 다음 기사：Python PDB 코드 검토 보고서
본 웹사이트의 성명
본 글의 내용은 네티즌들의 자발적인 기여로 작성되었으며, 저작권은 원저작자에게 있습니다. 본 사이트는 이에 상응하는 법적 책임을 지지 않습니다. 표절이나 침해가 의심되는 콘텐츠를 발견한 경우 admin@php.cn으로 문의하세요.
저자별 최신 기사
최신 이슈
function_exists()는 사용자 정의 함수를 결정할 수 없습니다. 기능 테스트() {                                                                                ...
에서 2024-04-29 11:01:01
0
2
1614
Chrome 모바일 버전을 표시하는 방법 안녕하세요 선생님, Chrome을 모바일 버전으로 어떻게 변경하나요?
에서 2024-04-23 00:22:19
0
10
1768
자식 창이 부모 창을 작동하지만 출력이 응답하지 않습니다. 처음 두 문장은 실행 가능하지만 마지막 문장은 구현할 수 없습니다.
에서 2024-04-19 15:37:47
0
1
1509
상위 창에 출력이 없습니다. document.onclick = function(){ window.opener.document.write('나는 자식 창의 출력입니다.');
에서 2024-04-18 23:52:34
0
1
1421
CSS 마인드맵 코스웨어는 어디에 있나요? 코스웨어
에서 2024-04-16 10:10:18
0
0
1466
관련 주제
더>
인기 추천
인기 튜토리얼
더>
관련 튜토리얼
인기 추천
최신 강좌
최신 다운로드
더>
웹 효과
웹사이트 소스 코드
웹사이트 자료
프론트엔드 템플릿
회사 소개 부인 성명 Sitemap
PHP 중국어 웹사이트：공공복지 온라인 PHP 교육，PHP 학습자의 빠른 성장을 도와주세요！