Easy Semantic Color (TF-IDF)

example:

https://gyazo.com/f5afd0e448225f72fc6b29aa1eefcdd7

1. 半角スペースで分かち書きされたデータを用意してください

code:sample.py

document_sample = [

"This is a test.",

"これ は サンプル です 。"

]

a. TFIDF

b. PCAで2次元に削減

c. 標準化

a. 極座標を計算

b. 中心角に色相環をあてはめ

結果の表示

code:color.py

from matplotlib import pyplot as plt

from sklearn.feature_extraction.text import TfidfVectorizer

from sklearn.decomposition import PCA

from sklearn.preprocessing import StandardScaler

import numpy as np

import matplotlib.colors as colors

def polar_coordinates_to_color(pca_result, saturation = 0.8, brightness = 0.8):

hues = []

for x, y in pca_result:

# Calculate the angle in radians and convert it to degrees

theta = np.arctan2(y, x)

theta_degree = np.degrees(theta) % 360

# Map the angle to the hue (color)

hues.append(hue)

# Convert to RGB colors

return rgb_colors

def perform_tfidf_pca_standardization(documents):

"""

Performs TF-IDF analysis on the provided documents, applies PCA to reduce to 2 dimensions,

and then standardizes the PCA result.

:param documents: A list of documents (strings)

:return: A numpy array representing the standardized 2D PCA result of the TF-IDF vectors

"""

# Step 1: TF-IDF Analysis

tfidf_vectorizer = TfidfVectorizer()

tfidf_matrix = tfidf_vectorizer.fit_transform(documents)

# Step 2: PCA to reduce to 2 dimensions

pca = PCA(n_components=2)

pca_result = pca.fit_transform(tfidf_matrix.toarray())

# Step 3: Standardize the PCA result

scaler = StandardScaler()

standardized_pca_result = scaler.fit_transform(pca_result)

return standardized_pca_result

def scatter_pca(pca_result):

# Convert PCA Results to Colors

colors_for_plotting = polar_coordinates_to_color(pca_result)

# Plotting the Scatter Plot

plt.figure(figsize=(8, 6))

plt.title("TF-IDF PCA Scatter Plot with Polar Coordinates Coloring")

plt.xlabel("PCA Dimension 1")

plt.ylabel("PCA Dimension 2")

plt.show()