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
# Ilosc wierszy 
NUMROWS=$(cat data.shuf | wc -l)
echo "Initial number of rows ${NUMROWS}"
echo '{print $1}'
CUTOFF=$1
DELETE=$((NUMROWS - CUTOFF))
echo "Cutting ${CUTOFF} rows"
tail -n $DELETE data.shuf > data.shuf.cut

# Podzial na 3 pozbiory
#Ustalenie proporcji 6:2:2
NUMROWS=$(cat data.shuf.cut | wc -l)
TEST=$((NUMROWS / 10 * 2 ))
DEV=$((NUMROWS / 10 * 4 ))
TRAIN=$((NUMROWS - DEV ))

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

rm data.shuf.cut

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