Korean, Edit

Python Syntax

Recommended Reading : 【Python】 Python Table of Contents

1. Overview

2. Syntax

3. Library

a. R Overview


1. Overview

⑴ Advantage 1. Easy to implement backend in app development, etc.

⑵ Advantage 2. Well-developed deep learning tools

⑶ Advantage 3. Various applications possible based on open community

⑷ Advantage 4. Excellent data visualization



2. Syntax

⑴ Python can return multiple values simultaneously

⑵ %%time: Outputs running time

⑶ pass or ...: Used when code is syntactically required but no action is needed.

⑷ yield returns values one at a time when needed, rather than all at once

⑸ You can write multi-line code by using backspace \ for line breaks

⑹ A single-line string is represented by “~” or ‘~’, and a multi-line string is represented by “””~””” or ‘’’~’’’

⑺ adata.var_names=[i.upper() for i in list(adata.var_names)]

① Meaning: In adata.var_names, each i in list(data.var_names) is stored in uppercase

⑻ if issparse(counts.X):counts.X=counts.X.A

① Meaning: If counts.X is a sparse matrix, the value of counts.X.A is assigned to counts.X

⑼ pd.DataFrame([(str(i),str(j)) for i in range(4) for j in range(i+1, 4)])


	0	1
0	0	1
1	0	2
2	0	3
3	1	2
4	1	3
5	2	3

⑽ Matrix Operations

① For matrices defined with NumPy, the * and ** operators perform element-wise multiplication (and element-wise power). → In this case, use the @ operator to indicate matrix multiplication.

② For matrices defined with SymPy (Matrix), the * and ** operators mean matrix multiplication and matrix power, respectively.

○ Using Python (NumPy)


import numpy as np

# Define two matrices
A = np.array([[1, 2], [3, 4]])
B = np.array([[5, 6], [7, 8]])

# Print matrices
print("Matrix A:")
print(A)
print("\nMatrix B:")
print(B)

# Matrix multiplication
result = np.dot(A, B)
## Alternatively, result = A @ B

print("\nNumPy matrix multiplication result:")
print(result)


○ Using Python (SymPy)


from sympy import Matrix, init_printing

# Initialize LaTeX-style printing
init_printing()

# Define two matrices
A = Matrix([[1, 2], [3, 4]])
B = Matrix([[5, 6], [7, 8]])

# Print matrices
print("Matrix A:")
display(A)

print("Matrix B:")
display(B)

# Matrix multiplication
result = A * B

print("SymPy matrix multiplication result:")
display(result)



3. Library

⑴ NumPy

① A core library that enables fast numerical computation (matrix operations, vector operations) based on multidimensional arrays (ndarray).

② Broadcasting

○ A feature in NumPy that allows operations between arrays of mismatched dimensions

○ A small array (as long as it has a dimension of 1) is automatically expanded to match the larger array’s shape, and then the operation is performed

○ Data is not physically copied to match the small array to the large array


# Shape of `exp_log_mu`: (2257, 6)
# Shape of `theta_mn`: (6, 13344)

exp_log_mu_expanded = exp_log_mu[:, :, np.newaxis]   # Shape: (2257, 6, 1)
theta_mn_expanded = theta_mn[np.newaxis, :, :]       # Shape: (1, 6, 13344)    
result = exp_log_mu_expanded * theta_mn_expanded     # Shape: (2257, 6, 13344)

⑵ Pandas

① Handles tabular data (DataFrame) easily and is strong for data cleaning/preprocessing (handling missing values, filtering, grouping).

⑶ Matplotlib

① Used to visualize data distributions or patterns by drawing graphs (line plots, bar charts, scatter plots, etc.).

⑷ Scikit-learn (skimage)

① Provides traditional machine learning algorithms (classification/regression/clustering) and evaluation tools (cross-validation, performance metrics).

⑸ TensorFlow

① Allows training and deployment of deep learning models using a computation-graph-based approach; used for large-scale training and production deployment.

② Developed by Google’s AI team for machine learning and deep learning purposes.

③ Released in November 2015.

⑹ PyTorch

① Based on a dynamic computation graph, making implementation intuitive; widely used in research/experimentation to quickly develop and train deep learning models.

② Developed by Facebook’s artificial intelligence research team.

③ First released in 2016 and stabilized in April 2019.

④ Used in various fields, including natural language processing.

⑤ Highly flexible, so it is widely used in fundamental and core technology research.

⑺ SciPy

① Offers advanced mathematical/scientific computing features (optimization, statistics, signal processing), used for numerical computation in modeling and analysis.

⑻ OpenCV

① An image/video processing library used for preprocessing before tasks like object detection (filtering, contour detection, feature extraction).

⑼ NLTK

① Provides basic NLP tools (tokenization, morphological analysis, corpus processing) and is commonly used for education and introductory practice.

⑽ Transformers (Hugging Face)

① Makes it easy to load pretrained language models like BERT and GPT and apply them to NLP tasks such as classification, summarization, and translation.

⑾ spaCy

① A fast, production-oriented NLP library that efficiently performs named entity recognition (NER), part-of-speech tagging, and dependency parsing.

⑿ JAX

① A library that enables high-performance numerical computing and deep learning in Python.

② JaxGCRL only supports older jaxlib versions.

③ Jax does not support GPU usage on Windows. (ref)


Input: 2024.04.30 13:41

results matching ""

    No results matching ""