Advanced Data Visualization
Interactive বাংলা লেকচার

# Colab cell 1: প্রয়োজনীয় লাইব্রেরি
!pip install wordcloud -q

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from wordcloud import WordCloud
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.decomposition import PCA
from sklearn.manifold import TSNE
from pandas.plotting import parallel_coordinates

sns.set_style("whitegrid")
plt.rcParams['figure.figsize'] = (8, 5)

np.random.seed(42)
n = 200
df = pd.DataFrame({
    "Age": np.random.randint(18, 65, n),
    "Annual_Income": np.random.randint(15, 140, n),
    "Spending_Score": np.random.randint(1, 100, n),
    "Gender": np.random.choice(["Male","Female"], n)
})
print(df.shape)
df.head()

reviews = [
    "the battery life is amazing and lasts long",
    "battery drains too fast very disappointing",
    "camera quality is amazing love the photos",
    "delivery was late but product quality is good",
    "battery heats up quickly during charging"
]
from collections import Counter
stop_words = {"the","is","was","and","but","too","up","during"}
all_words = []
for review in reviews:
    words = [w for w in review.lower().split() if w not in stop_words]
    all_words.extend(words)
word_freq = Counter(all_words).most_common(8)
words, counts = zip(*word_freq)
plt.barh(words, counts, color="#0F6E56")
plt.xlabel("Frequency"); plt.title("Top Words in Reviews")
plt.gca().invert_yaxis(); plt.show()

text_combined = " ".join(all_words)
wc = WordCloud(width=800, height=400, background_color="white",
               colormap="viridis").generate(text_combined)
plt.figure(figsize=(10,5))
plt.imshow(wc, interpolation="bilinear")
plt.axis("off"); plt.title("Customer Review Word Cloud"); plt.show()

vectorizer = TfidfVectorizer(stop_words="english")
tfidf_matrix = vectorizer.fit_transform(reviews)
feature_names = vectorizer.get_feature_names_out()
scores = tfidf_matrix.sum(axis=0).A1
tfidf_df = pd.DataFrame({"word": feature_names, "score": scores})
tfidf_df = tfidf_df.sort_values("score", ascending=False).head(10)
plt.barh(tfidf_df["word"], tfidf_df["score"], color="#993C1D")
plt.title("Top TF-IDF Words"); plt.gca().invert_yaxis(); plt.show()

corr = df[["Age","Annual_Income","Spending_Score"]].corr()
plt.figure()
sns.heatmap(corr, annot=True, cmap="coolwarm", vmin=-1, vmax=1)
plt.title("Feature Correlation Heatmap"); plt.show()

colors = df["Gender"].map({"Male":"#185FA5","Female":"#993C1D"})
plt.scatter(df["Annual_Income"], df["Spending_Score"],
            s=df["Age"]*3, c=colors, alpha=0.6)
plt.xlabel("Annual Income"); plt.ylabel("Spending Score")
plt.title("Income vs Spending (bubble size = Age)"); plt.show()

df_pc = df.copy()
df_pc["Segment"] = pd.qcut(df_pc["Spending_Score"], 2, labels=["Low","High"])
plt.figure(figsize=(9,5))
parallel_coordinates(df_pc[["Age","Annual_Income","Spending_Score","Segment"]],
                      "Segment", color=["#0F6E56","#993C1D"])
plt.title("Parallel Coordinates: Customer Segments"); plt.show()

months = ["Jan","Feb","Mar","Apr"]
sales_2024 = [120, 135, 150, 140]
sales_2025 = [160, 150, 170, 190]
x = np.arange(len(months)); width = 0.35
plt.bar(x - width/2, sales_2024, width, label="2024", color="#042C53")
plt.bar(x + width/2, sales_2025, width, label="2025", color="#993C1D")
plt.xticks(x, months); plt.legend(); plt.title("Sales: 2024 vs 2025"); plt.show()

from sklearn.datasets import load_iris
iris = load_iris(); X = iris.data; y = iris.target
pca = PCA(n_components=2); X_pca = pca.fit_transform(X)
plt.scatter(X_pca[:,0], X_pca[:,1], c=y, cmap="viridis")
plt.xlabel("PC1"); plt.ylabel("PC2")
plt.title("PCA: Iris (4D → 2D)"); plt.show()

tsne = TSNE(n_components=2, perplexity=30, random_state=42)
X_tsne = tsne.fit_transform(X)
plt.scatter(X_tsne[:,0], X_tsne[:,1], c=y, cmap="viridis")
plt.title("t-SNE: Iris Clusters"); plt.show()