ium_434784/preparations.sh

curl -OL https://git.wmi.amu.edu.pl/s434784/ium_434784/raw/branch/master/who_suicide_statistics.csv

# uciecie nazw kolumn 
head -n 1 who_suicide_statistics.csv > names.csv

# Randomizacja zbioru.
sed 1d who_suicide_statistics.csv | shuf > data.shuf
#cp who_suicide_statistics.csv data.shuf

# Podzial na 3 pozbiory

#Ustalenie proporcji 6:2:2
NUMROWS=$(cat data.shuf | wc -l)
TEST=$((NUMROWS / 10 * 2 ))
DEV=$((NUMROWS / 10 * 4 ))
TRAIN=$((NUMROWS - DEV ))

head -n $TEST data.shuf > data.test
head -n $DEV data.shuf | tail -n $TEST  > data.dev
head -n $NUMROWS data.shuf | tail -n $TRAIN > data.train

echo "Test rows ${TEST}"
echo "Dev rows ${DEV}"
echo "Train rows ${TRAIN}"
echo "All number of rows ${NUMROWS}"
echo "${1}"