s444018/ium_444018
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Zaktualizuj 'download.sh'

Browse Source
This commit is contained in:
Szymon Parafiński 2022-03-27 22:59:50 +02:00
parent 12fb47df5f
commit a98b89468b
1 changed files with 8 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
download.sh
View File

@ -1,15 +1,15 @@
dataset_operation() {
tail -n +2 all_games.csv | shuf > all_games.csv.s
head -n $CUTOFF all_games.csv.s > ./all_games.csv.shuf
len1=$(cat ./all_games.csv.shuf | wc -l)
tail -n +2 imdb_top_1000.csv | shuf > imdb_top_1000.csv.s
head -n $CUTOFF imdb_top_1000.csv.s > ./imdb_top_1000.csv.shuf
len1=$(cat ./imdb_top_1000.csv.shuf | wc -l)
len2=$(($len1/10))
len3=$(($len2*2))
len4=$(($len3+1))
head -n $len2 all_games.csv.shuf > all_games.csv.test
head -n $len3 all_games.csv.shuf | tail -n $len2 > all_games.csv.dev
tail -n +$len4 all_games.csv.shuf > all_games.csv.train
rm all_games.csv.shuf
wc -l all_games.csv.*
head -n $len2 imdb_top_1000.csv.shuf > imdb_top_1000.csv.test
head -n $len3 imdb_top_1000.csv.shuf | tail -n $len2 > imdb_top_1000.csv.dev
tail -n +$len4 imdb_top_1000.csv.shuf > imdb_top_1000.csv.train
rm imdb_top_1000.csv.shuf
wc -l imdb_top_1000.csv.*
}
kaggle datasets download -d harshitshankhdhar/imdb-dataset-of-top-1000-movies-and-tv-shows