ceshine

ceshine

from subprocess import Popen, PIPE


def image_markdown_conversion():
    try:
        while True:
            text = input("Input:").strip()
            brackets, split_point = 1, 0
            description = ""
            assert len(text) > 5, "wrong format!"

ceshine

import re


def cjk_detect(texts):
    # korean
    if re.search("[\uac00-\ud7a3]", texts):
        return "ko"
    # japanese
    if re.search("[\u3040-\u30ff]", texts):
        return "ja"

ceshine

from zipfile import ZipFile
import zipfile
from pathlib import Path
import base64
import sys
import io


def write_folder(zfile: ZipFile, dir_path: Path, prefix: str = ""):
    assert dir_path.is_dir()

ceshine

# WARNING: this script is out-dated since the last update of the Tourism Bureau website.
from pathlib import Path

import pandas as pd

SCHEMAS = [
    (201201, "schema/residence-2012-01.csv"),
    (201101, "schema/residence-2011-01.csv")
]
DATA_FILE_PATTERN = "raw_data/{year}-{month}.xls"

ceshine

Data Analysis Pieces

• Taiwan Tourism Dataset
  • An Analysis of the Impact of Chinese Tourism Ban to Taiwan
  • Country-level Analysis of Numbers of Visitors to Taiwan

Competition Summary / Notes

• Learning Notes from Planet Competition
• Kaggle Instacart Competition

ceshine

library(ggplot2)
library(ggthemes)

dat <- data.frame(
    name = c("CPU", "GPU", "TPU"), 
    time = c(3 * 3600 + 6 * 60 + 4, 3 * 60 + 16, 1 * 60 + 42)
)
dat$log_time = log(dat$time)

ggplot(data=dat, aes(x=name, y=log_time)) + 

ceshine

ceshine

ceshine