kaggle準備¶

steps¶

create account
login
run prepare-kaggle.sh

# prepare python environment
# kg command
pip install kaggle-cli

# warning: save raw password
# ~/.kaggle-cli/config
kg config -u username -p password -g

# for accept / python 3.6
pip install git+https://github.com/ksomemo/kaggle-accept

# make directory for competition
mkdir -p /path/to/kaggle-root/competition-name && cd $_
competition_name=`basename $(pwd)`

# accept
python -m kaggle_accept -c $competition_name

# download data
mkdir input && cd $_
kg download -c $competition_name && cd ..

mkdir notebooks
mkdir scripts

check data¶

# size
ls -lh ./input/*.zip
ls -lh ./input/*.csv

# n rows
wc -l ./input/*.csv

# n columns
for f in `ls ../input/*.csv`
do
    head -n 1 $f | awk -F, -v fname="$f" '{print NF "\t" fname}'
done