Skip to content

chinarobotlab/analysis_on_the_story_of_a_stone

 
 

Repository files navigation

Analysis on the Story of a Stone

Author: Yu Lou

Written for Python 3.

preprocess.py

Preprocess text.

Input

  • hlm.txt: original text

Output

  • preprocessing.txt: preprocessed text.

preprocess_chapters.py

Split text into chapters and preprocess them.

Input

  • hlm.txt: original text

Output

  • chapters(folder): preprocessed text. One file for each chapter, numbered from "1.text".

dict_creat.py

Creat dictionary.

Input

  • preprocessing.txt: preprocessed text.

Output

  • dict.csv: dictionary.

word_split.py

Split words apart.

Input

  • preprocessing.txt: preprocessed text.
  • dict.csv: dictionary.

Output

  • word_split.text: splitted text.

word_split_chapters.py

Split words apart in all chapters.

Input

  • preprocessing.txt: preprocessed text.
  • dict.csv: dictionary.
  • chapter(folder): preprocessed text for all chapters.

Output

  • chapter_split(folder): splitted text. One file for each chapter, numbered from "1.text".

word_count.py

Count words.

Input

  • word_split.text: splitted text.

Output

  • word_count.csv: counting result, sorted by number of occurence.

word_count_chapters.py

Count words in each chapters.

Input

  • chapter_split(folder): splitted text for each chapter.

Output

  • word_count_chapters.csv: counting result. One line per word and one chapter per column.

analysis.py

Do PCA analysis. Show result on screen.

Prerequisite

"sklearn", "numpy" and "matplotlib" is needed to run this program.

Input

  • word_count_chapters.csv: word counting result for each chapters.

Output

  • components.csv: weights for each components.

suffix_tree.py

Libary for suffix tree.

correctness_calculate.py

Calculate the correctness of word splitting algorithm.

Input

  • *_answer.txt: answer.
  • *_result.txt: result of the program.

("*" is file prefix)

About

Analysis on the novel "the Story of a Stone"

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%