[mlpack-svn] r15612 - in mlpack/conf/jenkins-conf/benchmark: . benchmark
fastlab-svn at coffeetalk-1.cc.gatech.edu
fastlab-svn at coffeetalk-1.cc.gatech.edu
Tue Aug 13 15:17:59 EDT 2013
Author: marcus
Date: Tue Aug 13 15:17:58 2013
New Revision: 15612
Log:
Adjust parameters for the range search method and clean the benchmark main method.
Modified:
mlpack/conf/jenkins-conf/benchmark/benchmark/run_benchmark.py
mlpack/conf/jenkins-conf/benchmark/config.yaml
Modified: mlpack/conf/jenkins-conf/benchmark/benchmark/run_benchmark.py
==============================================================================
--- mlpack/conf/jenkins-conf/benchmark/benchmark/run_benchmark.py (original)
+++ mlpack/conf/jenkins-conf/benchmark/benchmark/run_benchmark.py Tue Aug 13 15:17:58 2013
@@ -37,20 +37,6 @@
Log.Info('CPU Cores: ' + SystemInfo.GetCPUCores())
'''
-Check if the file is available in one of the given formats.
-
- at param dataset - Datsets which should be checked.
- at param formats - List of supported file formats.
- at return Orginal dataset or dataset with new file format.
-'''
-def CheckFileExtension(dataset, formats):
- dataExtension = os.path.splitext(dataset)[1][1:]
- if dataExtension in formats:
- return dataset
- else:
- return dataset[0:len(dataset) - len(dataExtension)] + formats[0]
-
-'''
Return a list with modified dataset.
@param dataset - Datasets to be modified.
@@ -92,19 +78,6 @@
return (datasetList, modifiedList)
'''
-This function Remove a given file or list of files.
-
- at param dataset - File or list of file which should be deleted.
-'''
-def RemoveDataset(dataset):
- if isinstance(dataset, str):
- dataset = [dataset]
-
- for f in dataset:
- if os.path.isfile(f):
- os.remove(f)
-
-'''
Count all datasets to determine the dataset size.
@param libraries - Contains the Dataset List.
@@ -138,7 +111,6 @@
streamData = config.StreamMerge()
# Read the general block and set the attributes.
-
if "general" in streamData:
for key, value in streamData["general"]:
if key == "timeout":
@@ -192,14 +164,12 @@
format = libary[4]
header.append(name)
-
-
if not blocks or name in blocks:
run += 1
Log.Info("Libary: " + name)
- # Logging: create a new build and libary record for this libary.
+ # Logging: create a new build and library record for this library.
if log and name not in build:
libaryId = db.GetLibrary(name)
libaryId = libaryId[0][0] if libaryId else db.NewLibrary(name)
@@ -216,21 +186,22 @@
else:
for dataset in datsets:
- datasetName = NormalizeDatasetName(dataset)
+ datasetName = NormalizeDatasetName(dataset)
row = FindRightRow(dataMatrix, datasetName, datasetCount)
# Logging: Create a new dataset record fot this dataset.
if log:
datasetId = db.GetDataset(datasetName)
- datasetId = datasetId[0][0] if datasetId else db.NewDataset(*DatasetInfo(dataset))
+ datasetId = datasetId[0][0] if datasetId else db.NewDataset(*DatasetInfo(dataset))
- dataMatrix[row][0] = NormalizeDatasetName(dataset)
- Log.Info("Dataset: " + dataMatrix[row][0])
+ dataMatrix[row][0] = datasetName
+ Log.Info("Dataset: " + dataMatrix[row][0])
modifiedDataset = GetDataset(dataset, format)
try:
- instance = methodCall(modifiedDataset[0], timeout=timeout, verbose=False)
+ instance = methodCall(modifiedDataset[0], timeout=timeout,
+ verbose=False)
except Exception as e:
Log.Fatal("Could not call the constructor: " + script)
Log.Fatal("Exception: " + str(e))
@@ -265,7 +236,8 @@
var = sum((avg - value) ** 2 for value in time) / len(time)
buildId, libaryId = build[name]
- db.NewResult(buildId, libaryId, dataMatrix[row][col], var, datasetId, methodId)
+ db.NewResult(buildId, libaryId, dataMatrix[row][col], var,
+ datasetId, methodId)
# Remove temporary datasets.
RemoveDataset(modifiedDataset[1])
Modified: mlpack/conf/jenkins-conf/benchmark/config.yaml
==============================================================================
--- mlpack/conf/jenkins-conf/benchmark/config.yaml (original)
+++ mlpack/conf/jenkins-conf/benchmark/config.yaml Tue Aug 13 15:17:58 2013
@@ -88,50 +88,50 @@
datasets:
- files: ['datasets/circle_data.csv', 'datasets/stock.csv',
'datasets/abalone.csv', 'datasets/bank8FM.csv',
- 'datasets/waveform.csv', 'datasets/cal_housing.csv',
- 'datasets/TomsHardware.csv', 'datasets/arcene_X.csv',
- 'datasets/madelon_X.csv', 'datasets/pendigits.csv',
- 'datasets/isolet.csv', 'datasets/covtype.csv']
+ 'datasets/waveform.csv', 'datasets/TomsHardware.csv',
+ 'datasets/arcene_X.csv', 'datasets/madelon_X.csv',
+ 'datasets/pendigits.csv', 'datasets/isolet.csv',
+ 'datasets/covtype.csv']
options: '-k linear'
- files: ['datasets/circle_data.csv', 'datasets/stock.csv',
'datasets/abalone.csv', 'datasets/bank8FM.csv',
- 'datasets/waveform.csv', 'datasets/cal_housing.csv',
- 'datasets/TomsHardware.csv', 'datasets/arcene_X.csv',
- 'datasets/madelon_X.csv', 'datasets/pendigits.csv',
- 'datasets/isolet.csv', 'datasets/covtype.csv']
+ 'datasets/waveform.csv', 'datasets/TomsHardware.csv',
+ 'datasets/arcene_X.csv', 'datasets/madelon_X.csv',
+ 'datasets/pendigits.csv', 'datasets/isolet.csv',
+ 'datasets/covtype.csv']
options: '-k gaussian'
- files: ['datasets/circle_data.csv', 'datasets/stock.csv',
'datasets/abalone.csv', 'datasets/bank8FM.csv',
- 'datasets/waveform.csv', 'datasets/cal_housing.csv',
- 'datasets/TomsHardware.csv', 'datasets/arcene_X.csv',
- 'datasets/madelon_X.csv', 'datasets/pendigits.csv',
- 'datasets/isolet.csv', 'datasets/covtype.csv']
+ 'datasets/waveform.csv', 'datasets/TomsHardware.csv',
+ 'datasets/arcene_X.csv', 'datasets/madelon_X.csv',
+ 'datasets/pendigits.csv', 'datasets/isolet.csv',
+ 'datasets/covtype.csv']
options: '-k polynomial'
- files: ['datasets/circle_data.csv', 'datasets/stock.csv',
'datasets/abalone.csv', 'datasets/bank8FM.csv',
- 'datasets/waveform.csv', 'datasets/cal_housing.csv',
- 'datasets/TomsHardware.csv', 'datasets/arcene_X.csv',
- 'datasets/madelon_X.csv', 'datasets/pendigits.csv',
- 'datasets/isolet.csv', 'datasets/covtype.csv']
+ 'datasets/waveform.csv', 'datasets/TomsHardware.csv',
+ 'datasets/arcene_X.csv', 'datasets/madelon_X.csv',
+ 'datasets/pendigits.csv', 'datasets/isolet.csv',
+ 'datasets/covtype.csv']
options: '-k hyptan'
- files: ['datasets/circle_data.csv', 'datasets/stock.csv',
'datasets/abalone.csv', 'datasets/bank8FM.csv',
- 'datasets/waveform.csv', 'datasets/cal_housing.csv',
- 'datasets/TomsHardware.csv', 'datasets/arcene_X.csv',
- 'datasets/madelon_X.csv', 'datasets/pendigits.csv',
- 'datasets/isolet.csv', 'datasets/covtype.csv']
+ 'datasets/waveform.csv', 'datasets/TomsHardware.csv',
+ 'datasets/arcene_X.csv', 'datasets/madelon_X.csv',
+ 'datasets/pendigits.csv', 'datasets/isolet.csv',
+ 'datasets/covtype.csv']
options: '-k laplacian'
- files: ['datasets/circle_data.csv', 'datasets/stock.csv',
'datasets/abalone.csv', 'datasets/bank8FM.csv',
- 'datasets/waveform.csv', 'datasets/cal_housing.csv',
- 'datasets/TomsHardware.csv', 'datasets/arcene_X.csv',
- 'datasets/madelon_X.csv', 'datasets/pendigits.csv',
- 'datasets/isolet.csv', 'datasets/covtype.csv']
+ 'datasets/waveform.csv', 'datasets/TomsHardware.csv',
+ 'datasets/arcene_X.csv', 'datasets/madelon_X.csv',
+ 'datasets/pendigits.csv', 'datasets/isolet.csv',
+ 'datasets/covtype.csv']
options: '-k cosine'
LARS:
run: true
@@ -303,21 +303,21 @@
'datasets/madelon_X.csv', 'datasets/arcene_X.csv',
'datasets/corel-histogram.csv', 'datasets/isolet.csv',
'datasets/covtype.csv', 'datasets/Twitter.csv']
- options: '-M 20.3'
+ options: '-M 0.02'
- files: ['datasets/wine.csv', 'datasets/ionosphere.csv',
'datasets/cloud.csv', 'datasets/vehicle.csv',
'datasets/madelon_X.csv', 'datasets/arcene_X.csv',
'datasets/corel-histogram.csv', 'datasets/isolet.csv',
'datasets/covtype.csv', 'datasets/Twitter.csv']
- options: '-M 20.3 -N'
+ options: '-M 0.02 -N'
- files: ['datasets/wine.csv', 'datasets/ionosphere.csv',
'datasets/cloud.csv', 'datasets/vehicle.csv',
'datasets/madelon_X.csv', 'datasets/arcene_X.csv',
'datasets/corel-histogram.csv', 'datasets/isolet.csv',
'datasets/covtype.csv', 'datasets/Twitter.csv']
- options: '-M 20.3 -s'
+ options: '-M 0.02 -s'
GMM:
run: true
script: methods/mlpack/gmm.py
@@ -651,7 +651,7 @@
'datasets/madelon_X.csv', 'datasets/arcene_X.csv',
'datasets/corel-histogram.csv', 'datasets/isolet.csv',
'datasets/covtype.csv', 'datasets/Twitter.csv']
- options: '-M 20.3'
+ options: '-M 0.02'
LinearRegression:
run: true
iteration: 3
@@ -781,26 +781,26 @@
datasets:
- files: ['datasets/circle_data.csv', 'datasets/stock.csv',
'datasets/abalone.csv', 'datasets/bank8FM.csv',
- 'datasets/waveform.csv', 'datasets/cal_housing.csv',
- 'datasets/TomsHardware.csv', 'datasets/arcene_X.csv',
- 'datasets/madelon_X.csv', 'datasets/pendigits.csv',
- 'datasets/isolet.csv', 'datasets/covtype.csv']
+ 'datasets/waveform.csv', 'datasets/TomsHardware.csv',
+ 'datasets/arcene_X.csv', 'datasets/madelon_X.csv',
+ 'datasets/pendigits.csv', 'datasets/isolet.csv',
+ 'datasets/covtype.csv']
options: '-k linear'
- files: ['datasets/circle_data.csv', 'datasets/stock.csv',
'datasets/abalone.csv', 'datasets/bank8FM.csv',
- 'datasets/waveform.csv', 'datasets/cal_housing.csv',
- 'datasets/TomsHardware.csv', 'datasets/arcene_X.csv',
- 'datasets/madelon_X.csv', 'datasets/pendigits.csv',
- 'datasets/isolet.csv', 'datasets/covtype.csv']
+ 'datasets/waveform.csv', 'datasets/TomsHardware.csv',
+ 'datasets/arcene_X.csv', 'datasets/madelon_X.csv',
+ 'datasets/pendigits.csv', 'datasets/isolet.csv',
+ 'datasets/covtype.csv']
options: '-k polynomial'
- files: ['datasets/circle_data.csv', 'datasets/stock.csv',
'datasets/abalone.csv', 'datasets/bank8FM.csv',
- 'datasets/waveform.csv', 'datasets/cal_housing.csv',
- 'datasets/TomsHardware.csv', 'datasets/arcene_X.csv',
- 'datasets/madelon_X.csv', 'datasets/pendigits.csv',
- 'datasets/isolet.csv', 'datasets/covtype.csv']
+ 'datasets/waveform.csv', 'datasets/TomsHardware.csv',
+ 'datasets/arcene_X.csv', 'datasets/madelon_X.csv',
+ 'datasets/pendigits.csv', 'datasets/isolet.csv',
+ 'datasets/covtype.csv']
options: '-k hyptan'
LARS:
run: true
@@ -943,26 +943,26 @@
datasets:
- files: ['datasets/circle_data.csv', 'datasets/stock.csv',
'datasets/abalone.csv', 'datasets/bank8FM.csv',
- 'datasets/waveform.csv', 'datasets/cal_housing.csv',
- 'datasets/TomsHardware.csv', 'datasets/arcene_X.csv',
- 'datasets/madelon_X.csv', 'datasets/pendigits.csv',
- 'datasets/isolet.csv', 'datasets/covtype.csv']
+ 'datasets/waveform.csv', 'datasets/TomsHardware.csv',
+ 'datasets/arcene_X.csv', 'datasets/madelon_X.csv',
+ 'datasets/pendigits.csv', 'datasets/isolet.csv',
+ 'datasets/covtype.csv']
options: '-k linear'
- files: ['datasets/circle_data.csv', 'datasets/stock.csv',
'datasets/abalone.csv', 'datasets/bank8FM.csv',
- 'datasets/waveform.csv', 'datasets/cal_housing.csv',
- 'datasets/TomsHardware.csv', 'datasets/arcene_X.csv',
- 'datasets/madelon_X.csv', 'datasets/pendigits.csv',
- 'datasets/isolet.csv', 'datasets/covtype.csv']
+ 'datasets/waveform.csv', 'datasets/TomsHardware.csv',
+ 'datasets/arcene_X.csv', 'datasets/madelon_X.csv',
+ 'datasets/pendigits.csv', 'datasets/isolet.csv',
+ 'datasets/covtype.csv']
options: '-k gaussian'
- files: ['datasets/circle_data.csv', 'datasets/stock.csv',
'datasets/abalone.csv', 'datasets/bank8FM.csv',
- 'datasets/waveform.csv', 'datasets/cal_housing.csv',
- 'datasets/TomsHardware.csv', 'datasets/arcene_X.csv',
- 'datasets/madelon_X.csv', 'datasets/pendigits.csv',
- 'datasets/isolet.csv', 'datasets/covtype.csv']
+ 'datasets/waveform.csv', 'datasets/TomsHardware.csv',
+ 'datasets/arcene_X.csv', 'datasets/madelon_X.csv',
+ 'datasets/pendigits.csv', 'datasets/isolet.csv',
+ 'datasets/covtype.csv']
options: '-k hyptan'
LARS:
run: true
@@ -1080,34 +1080,34 @@
datasets:
- files: ['datasets/circle_data.csv', 'datasets/stock.csv',
'datasets/abalone.csv', 'datasets/bank8FM.csv',
- 'datasets/waveform.csv', 'datasets/cal_housing.csv',
- 'datasets/TomsHardware.csv', 'datasets/arcene_X.csv',
- 'datasets/madelon_X.csv', 'datasets/pendigits.csv',
- 'datasets/isolet.csv', 'datasets/covtype.csv']
+ 'datasets/waveform.csv', 'datasets/TomsHardware.csv',
+ 'datasets/arcene_X.csv', 'datasets/madelon_X.csv',
+ 'datasets/pendigits.csv', 'datasets/isolet.csv',
+ 'datasets/covtype.csv']
options: '-k linear'
- files: ['datasets/circle_data.csv', 'datasets/stock.csv',
'datasets/abalone.csv', 'datasets/bank8FM.csv',
- 'datasets/waveform.csv', 'datasets/cal_housing.csv',
- 'datasets/TomsHardware.csv', 'datasets/arcene_X.csv',
- 'datasets/madelon_X.csv', 'datasets/pendigits.csv',
- 'datasets/isolet.csv', 'datasets/covtype.csv']
+ 'datasets/waveform.csv', 'datasets/TomsHardware.csv',
+ 'datasets/arcene_X.csv', 'datasets/madelon_X.csv',
+ 'datasets/pendigits.csv', 'datasets/isolet.csv',
+ 'datasets/covtype.csv']
options: '-k gaussian'
- files: ['datasets/circle_data.csv', 'datasets/stock.csv',
'datasets/abalone.csv', 'datasets/bank8FM.csv',
- 'datasets/waveform.csv', 'datasets/cal_housing.csv',
- 'datasets/TomsHardware.csv', 'datasets/arcene_X.csv',
- 'datasets/madelon_X.csv', 'datasets/pendigits.csv',
- 'datasets/isolet.csv', 'datasets/covtype.csv']
+ 'datasets/waveform.csv', 'datasets/TomsHardware.csv',
+ 'datasets/arcene_X.csv', 'datasets/madelon_X.csv',
+ 'datasets/pendigits.csv', 'datasets/isolet.csv',
+ 'datasets/covtype.csv']
options: '-k polynomial'
- files: ['datasets/circle_data.csv', 'datasets/stock.csv',
'datasets/abalone.csv', 'datasets/bank8FM.csv',
- 'datasets/waveform.csv', 'datasets/cal_housing.csv',
- 'datasets/TomsHardware.csv', 'datasets/arcene_X.csv',
- 'datasets/madelon_X.csv', 'datasets/pendigits.csv',
- 'datasets/isolet.csv', 'datasets/covtype.csv']
+ 'datasets/waveform.csv', 'datasets/TomsHardware.csv',
+ 'datasets/arcene_X.csv', 'datasets/madelon_X.csv',
+ 'datasets/pendigits.csv', 'datasets/isolet.csv',
+ 'datasets/covtype.csv']
options: '-k hyptan'
NBC:
run: true
More information about the mlpack-svn
mailing list