Group CLI demo into subdirectory. (#6258)

CLI is not most developed interface. Putting them into correct directory can help new users to avoid it as most of the use cases are from a language binding.

demo/CLI/binary_classification/mushroom.conf

@@ -18,7 +18,7 @@ max_depth = 3
 # the number of round to do boosting
 num_round = 2
 # 0 means do not save any model except the final round model
-save_period = 0
+save_period = 2
 # The path of training data
 data = "agaricus.txt.train"
 # The path of validation data, used to monitor training process, here [test] sets name of the validation set

demo/CLI/binary_classification/runexp.sh

@@ -3,13 +3,15 @@
 python mapfeat.py
 # split train and test
 python mknfold.py agaricus.txt 1
-# training and output the models
-../../xgboost mushroom.conf
-# output prediction task=pred 
-../../xgboost mushroom.conf task=pred model_in=0002.model
-# print the boosters of 00002.model in dump.raw.txt
-../../xgboost mushroom.conf task=dump model_in=0002.model name_dump=dump.raw.txt 
-# use the feature map in printing for better visualization
-../../xgboost mushroom.conf task=dump model_in=0002.model fmap=featmap.txt name_dump=dump.nice.txt
-cat dump.nice.txt
 
+XGBOOST=../../../xgboost
+
+# training and output the models
+$XGBOOST mushroom.conf
+# output prediction task=pred
+$XGBOOST mushroom.conf task=pred model_in=0002.model
+# print the boosters of 00002.model in dump.raw.txt
+$XGBOOST mushroom.conf task=dump model_in=0002.model name_dump=dump.raw.txt
+# use the feature map in printing for better visualization
+$XGBOOST mushroom.conf task=dump model_in=0002.model fmap=featmap.txt name_dump=dump.nice.txt
+cat dump.nice.txt

demo/CLI/distributed-training/run_aws.sh

@@ -0,0 +1,11 @@
+# This is the example script to run distributed xgboost on AWS.
+# Change the following two lines for configuration
+
+export BUCKET=mybucket
+
+# submit the job to YARN
+../../../dmlc-core/tracker/dmlc-submit --cluster=yarn --num-workers=2 --worker-cores=2\
+				       ../../../xgboost mushroom.aws.conf nthread=2\
+				       data=s3://${BUCKET}/xgb-demo/train\
+				       eval[test]=s3://${BUCKET}/xgb-demo/test\
+				       model_dir=s3://${BUCKET}/xgb-demo/model

demo/CLI/regression/mapfeat.py

@@ -0,0 +1,33 @@
+#!/usr/bin/python
+
+fo = open('machine.txt', 'w')
+cnt = 6
+fmap = {}
+for l in open('machine.data'):
+    arr = l.split(',')
+    fo.write(arr[8])
+    for i in range(0, 6):
+        fo.write(' %d:%s' % (i, arr[i + 2]))
+
+    if arr[0] not in fmap:
+        fmap[arr[0]] = cnt
+        cnt += 1
+
+    fo.write(' %d:1' % fmap[arr[0]])
+    fo.write('\n')
+
+fo.close()
+
+# create feature map for machine data
+fo = open('featmap.txt', 'w')
+# list from machine.names
+names = [
+    'vendor', 'MYCT', 'MMIN', 'MMAX', 'CACH', 'CHMIN', 'CHMAX', 'PRP', 'ERP'
+]
+
+for i in range(0, 6):
+    fo.write('%d\t%s\tint\n' % (i, names[i + 1]))
+
+for v, k in sorted(fmap.items(), key=lambda x: x[1]):
+    fo.write('%d\tvendor=%s\ti\n' % (k, v))
+fo.close()

demo/CLI/regression/mknfold.py

@@ -0,0 +1,28 @@
+#!/usr/bin/python
+import sys
+import random
+
+if len(sys.argv) < 2:
+    print('Usage:<filename> <k> [nfold = 5]')
+    exit(0)
+
+random.seed(10)
+
+k = int(sys.argv[2])
+if len(sys.argv) > 3:
+    nfold = int(sys.argv[3])
+else:
+    nfold = 5
+
+fi = open(sys.argv[1], 'r')
+ftr = open(sys.argv[1] + '.train', 'w')
+fte = open(sys.argv[1] + '.test', 'w')
+for l in fi:
+    if random.randint(1, nfold) == k:
+        fte.write(l)
+    else:
+        ftr.write(l)
+
+fi.close()
+ftr.close()
+fte.close()

demo/CLI/yearpredMSD/csv2libsvm.py

@@ -1,14 +1,9 @@
-#!/usr/bin/python
 import sys
-
-if len(sys.argv) < 3:
-    print 'Usage: <csv> <libsvm>'
-    print 'convert a all numerical csv to libsvm'
-
 fo = open(sys.argv[2], 'w')
+
 for l in open(sys.argv[1]):
     arr = l.split(',')
     fo.write('%s' % arr[0])
-    for i in xrange(len(arr) - 1):
+    for i in range(len(arr) - 1):
         fo.write(' %d:%s' % (i, arr[i+1]))
 fo.close()

demo/CLI/yearpredMSD/runexp.sh

@@ -14,4 +14,4 @@ python csv2libsvm.py YearPredictionMSD.txt yearpredMSD.libsvm
 head -n 463715 yearpredMSD.libsvm > yearpredMSD.libsvm.train
 tail -n 51630 yearpredMSD.libsvm > yearpredMSD.libsvm.test
 echo "finish making the data"
-../../xgboost yearpredMSD.conf
+../../../xgboost yearpredMSD.conf

demo/distributed-training/run_aws.sh

@@ -1,11 +0,0 @@
-# This is the example script to run distributed xgboost on AWS.
-# Change the following two lines for configuration
-
-export BUCKET=mybucket
-
-# submit the job to YARN
-../../dmlc-core/tracker/dmlc-submit --cluster=yarn --num-workers=2 --worker-cores=2\
-    ../../xgboost mushroom.aws.conf nthread=2\
-    data=s3://${BUCKET}/xgb-demo/train\
-    eval[test]=s3://${BUCKET}/xgb-demo/test\
-    model_dir=s3://${BUCKET}/xgb-demo/model

demo/rank/runexp.sh

@@ -1,5 +1,4 @@
 #!/bin/bash
 
 ../../xgboost mq2008.conf
-
 ../../xgboost mq2008.conf task=pred model_in=0004.model

demo/rank/trans_data.py

@@ -38,4 +38,3 @@ if __name__ == "__main__":
     fi.close()
     output_feature.close()
     output_group.close()
-

demo/rank/wgetdata.sh

@@ -1,7 +1,13 @@
 #!/bin/bash
+if [ -f MQ2008.rar ]
+then
+    echo "Use downloaded data to run experiment."
+else
+    echo "Downloading data."
     wget https://s3-us-west-2.amazonaws.com/xgboost-examples/MQ2008.rar
     unrar x MQ2008.rar
     mv -f MQ2008/Fold1/*.txt .
+fi
 
 python trans_data.py train.txt mq2008.train mq2008.train.group
 

demo/regression/mapfeat.py

@@ -1,31 +0,0 @@
-#!/usr/bin/python
-
-fo = open( 'machine.txt', 'w' )
-cnt = 6
-fmap = {}
-for l in open( 'machine.data' ):
-    arr = l.split(',')
-    fo.write(arr[8])
-    for i in range( 0,6 ):
-        fo.write( ' %d:%s' %(i,arr[i+2]) )
-
-    if arr[0] not in fmap:
-        fmap[arr[0]] = cnt
-        cnt += 1
-
-    fo.write( ' %d:1' % fmap[arr[0]] )
-    fo.write('\n')
-
-fo.close()
-
-# create feature map for machine data
-fo = open('featmap.txt', 'w')
-# list from machine.names
-names = ['vendor','MYCT', 'MMIN', 'MMAX', 'CACH', 'CHMIN', 'CHMAX', 'PRP', 'ERP' ];
-
-for i in range(0,6):
-    fo.write( '%d\t%s\tint\n' % (i, names[i+1]))
-
-for v, k in sorted( fmap.items(), key = lambda x:x[1] ):
-    fo.write( '%d\tvendor=%s\ti\n' % (k, v))
-fo.close()

demo/regression/mknfold.py

@@ -1,29 +0,0 @@
-#!/usr/bin/python
-import sys
-import random
-
-if len(sys.argv) < 2:
-    print ('Usage:<filename> <k> [nfold = 5]')
-    exit(0)
-
-random.seed( 10 )
-
-k = int( sys.argv[2] )
-if len(sys.argv) > 3:
-    nfold = int( sys.argv[3] )
-else:
-    nfold = 5
-
-fi = open( sys.argv[1], 'r' )
-ftr = open( sys.argv[1]+'.train', 'w' )
-fte = open( sys.argv[1]+'.test', 'w' )
-for l in fi:
-    if random.randint( 1 , nfold ) == k:
-        fte.write( l )
-    else:
-        ftr.write( l )
-
-fi.close()
-ftr.close()
-fte.close()
-

jvm-packages/create_jni.py

@@ -116,7 +116,7 @@ if __name__ == "__main__":
             run("cmake .. " + " ".join(args) + maybe_generator)
             run("cmake --build . --config Release" + maybe_parallel_build)
 
-        with cd("demo/regression"):
+        with cd("demo/CLI/regression"):
             run(sys.executable + " mapfeat.py")
             run(sys.executable + " mknfold.py machine.txt 1")
 
@@ -138,11 +138,11 @@ if __name__ == "__main__":
 
     print("copying train/test files")
     maybe_makedirs("{}/src/test/resources".format(xgboost4j_spark))
-    with cd("../demo/regression"):
+    with cd("../demo/CLI/regression"):
         run("{} mapfeat.py".format(sys.executable))
         run("{} mknfold.py machine.txt 1".format(sys.executable))
 
-    for file in glob.glob("../demo/regression/machine.txt.t*"):
+    for file in glob.glob("../demo/CLI/regression/machine.txt.t*"):
         cp(file, "{}/src/test/resources".format(xgboost4j_spark))
     for file in glob.glob("../demo/data/agaricus.*"):
         cp(file, "{}/src/test/resources".format(xgboost4j_spark))

tests/python/test_demos.py

@@ -2,11 +2,13 @@ import os
 import subprocess
 import pytest
 import testing as tm
+import sys
 
 
 ROOT_DIR = tm.PROJECT_ROOT
 DEMO_DIR = os.path.join(ROOT_DIR, 'demo')
 PYTHON_DEMO_DIR = os.path.join(DEMO_DIR, 'guide-python')
+CLI_DEMO_DIR = os.path.join(DEMO_DIR, 'CLI')
 
 
 def test_basic_walkthrough():
@@ -132,7 +134,7 @@ def test_callbacks_demo():
 
 
 def test_cli_regression_demo():
-    reg_dir = os.path.join(DEMO_DIR, 'regression')
+    reg_dir = os.path.join(CLI_DEMO_DIR, 'regression')
     script = os.path.join(reg_dir, 'mapfeat.py')
     cmd = ['python', script]
     subprocess.check_call(cmd, cwd=reg_dir)
@@ -144,3 +146,15 @@ def test_cli_regression_demo():
     exe = os.path.join(tm.PROJECT_ROOT, 'xgboost')
     conf = os.path.join(reg_dir, 'machine.conf')
     subprocess.check_call([exe, conf], cwd=reg_dir)
+
+
+@pytest.mark.skipif(condition=sys.platform.startswith("win"),
+                    reason='Test requires sh execution.')
+def test_cli_binary_classification():
+    cls_dir = os.path.join(CLI_DEMO_DIR, 'binary_classification')
+    with tm.DirectoryExcursion(cls_dir, cleanup=True):
+        subprocess.check_call(['./runexp.sh'])
+        os.remove('0002.model')
+
+# year prediction is not tested due to data size being too large.
+# rank is not tested as it requires unrar command.

tests/python/testing.py

@@ -251,6 +251,36 @@ def eval_error_metric(predt, dtrain: xgb.DMatrix):
     return 'CustomErr', np.sum(r)
 
 
+class DirectoryExcursion:
+    def __init__(self, path: os.PathLike, cleanup=False):
+        '''Change directory.  Change back and optionally cleaning up the directory when exit.
+
+        '''
+        self.path = path
+        self.curdir = os.path.normpath(os.path.abspath(os.path.curdir))
+        self.cleanup = cleanup
+        self.files = {}
+
+    def __enter__(self):
+        os.chdir(self.path)
+        if self.cleanup:
+            self.files = {
+                os.path.join(root, f)
+                for root, subdir, files in os.walk(self.path) for f in files
+            }
+
+    def __exit__(self, *args):
+        os.chdir(self.curdir)
+        if self.cleanup:
+            files = {
+                os.path.join(root, f)
+                for root, subdir, files in os.walk(self.path) for f in files
+            }
+            diff = files.difference(self.files)
+            for f in diff:
+                os.remove(f)
+
+
 CURDIR = os.path.normpath(os.path.abspath(os.path.dirname(__file__)))
 PROJECT_ROOT = os.path.normpath(
     os.path.join(CURDIR, os.path.pardir, os.path.pardir))