iris-dataset-structures

Playing around with Datasets & Python w/ Copilot

🌟 A Tiny “Iris Dataset Quick Reference” You Can Keep Handy

🔧 Load the dataset

from sklearn.datasets import load_iris
import pandas as pd

iris = load_iris()
df = pd.DataFrame(iris.data, columns=iris.feature_names)
df["class"] = iris.target

---

📘 Core DataFrame operations

• df.head() — first rows
• df.tail() — last rows
• df.shape — dimensions
• df.describe() — stats summary
• df.columns — column names
• df.dtypes — data types

---

🔍 Exploration

• df["class"].unique() — class labels
• df["class"].value_counts() — samples per class
• df.corr() — correlations

---

📊 Quick visualizations

df.hist()
df.plot(kind="box")
df.plot(kind="scatter", x="sepal length (cm)", y="petal width (cm)")

---

🤖 Machine learning basics

from sklearn.model_selection import train_test_split
from sklearn.neighbors import KNeighborsClassifier

X = df.iloc[:, :-1]
y = df["class"]

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)

model = KNeighborsClassifier()
model.fit(X_train, y_train)
model.score(X_test, y_test)

#-----------------------------------------------

Below is a compact but powerful toolbox you can lean on as you learn.

---

🌱 Essential Iris Dataset Functions (Modern Python)

These examples assume:

from sklearn.datasets import load_iris
import pandas as pd

iris = load_iris()
df = pd.DataFrame(iris.data, columns=iris.feature_names)
df["class"] = iris.target

---

📌 1. Inspecting the Dataset

View the first rows

df.head()

View the last rows

df.tail()

Get dataset shape

df.shape

Column names

df.columns

Quick summary statistics

df.describe()

Data types

df.dtypes

---

🔍 2. Exploring the Data

Unique class labels

df["class"].unique()

Count samples per class

df["class"].value_counts()

Correlation matrix

df.corr()

---

📊 3. Basic Visualizations (using pandas + matplotlib)

import matplotlib.pyplot as plt

Histogram

df.hist(figsize=(8,6))
plt.show()

Scatter plot

df.plot(kind="scatter", x="sepal length (cm)", y="sepal width (cm)")
plt.show()

Boxplots

df.plot(kind="box", figsize=(8,6))
plt.show()

---

🤖 4. Preparing Data for Machine Learning

Split into train/test

from sklearn.model_selection import train_test_split

X = df.iloc[:, :-1]
y = df["class"]

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)

Train a simple model

from sklearn.neighbors import KNeighborsClassifier

model = KNeighborsClassifier()
model.fit(X_train, y_train)

Evaluate

model.score(X_test, y_test)

---

🧭 5. Converting the scikit‑learn dataset into a DataFrame (the clean way)

This is the modern, canonical approach:

df = pd.DataFrame(iris.data, columns=iris.feature_names)
df["class"] = iris.target

No URLs, no CSVs — just clean, in‑memory data.

---