add CUTOFF parameter

This commit is contained in:
Michał Dudziak 2023-03-23 12:28:29 +01:00
parent ee709e639c
commit f5cc717ffb

14
Jenkinsfile vendored
View File

@ -12,7 +12,12 @@ node {
defaultValue: '', defaultValue: '',
description: 'Kaggle token taken from kaggle.json file, as described in https://github.com/Kaggle/kaggle-api#api-credentials', description: 'Kaggle token taken from kaggle.json file, as described in https://github.com/Kaggle/kaggle-api#api-credentials',
name: 'KAGGLE_KEY' name: 'KAGGLE_KEY'
) ),
string(
defaultValue: '117926',
description: 'Number of positions to be saved',
name: 'CUTOFF',
trim: false)
]) ])
]) ])
} }
@ -28,12 +33,13 @@ node {
checkout scmGit(branches: [[name: '*/master']], extensions: [], userRemoteConfigs: [[url: 'https://git.wmi.amu.edu.pl/s452662/ium_452662']]) checkout scmGit(branches: [[name: '*/master']], extensions: [], userRemoteConfigs: [[url: 'https://git.wmi.amu.edu.pl/s452662/ium_452662']])
} }
stage('Cut data') { stage('Cut data') {
withEnv(["CUTOFF=${params.CUTOFF}"]) {
sh '''kaggle datasets download -d aleksandrglotov/car-prices-poland sh '''kaggle datasets download -d aleksandrglotov/car-prices-poland
unzip -o car-prices-poland.zip unzip -o car-prices-poland.zip
wc -l Car_Prices_Poland_Kaggle.csv wc -l Car_Prices_Poland_Kaggle.csv
head -n 11793 Car_Prices_Poland_Kaggle.csv > cars_test.csv head -n $(($CUTTOFF / 10)) Car_Prices_Poland_Kaggle.csv > cars_test.csv
head -n 23586 Car_Prices_Poland_Kaggle.csv | tail -11793 > cars_dev.csv head -n $(($CUTTOFF / 5)) Car_Prices_Poland_Kaggle.csv | tail $(($CUTTOFF / -10)) > cars_dev.csv
head -n 117926 Car_Prices_Poland_Kaggle.csv | tail -94340 > cars_train.csv''' head -n $(($CUTTOFF) Car_Prices_Poland_Kaggle.csv | tail $(($CUTTOFF * (-4 / 5) )) > cars_train.csv'''}
} }
stage('Save data into artifacts') { stage('Save data into artifacts') {
archiveArtifacts artifacts: '*.csv', followSymlinks: false archiveArtifacts artifacts: '*.csv', followSymlinks: false