Skip to content
/ daal Public
forked from oneapi-src/oneDAL

Intel® Data Analytics Acceleration Library (Intel® DAAL)

License

Notifications You must be signed in to change notification settings

vepifanov/daal

 
 

Repository files navigation

Intel® oneAPI Data Analytics Library

Installation   |   Documentation   |   Examples   |   Get Help   |   How to Contribute   

Build Status License

Intel® oneAPI Data Analytics Library (oneDAL) is a powerful machine learning library that helps speed up big data analysis. oneDAL solvers are also used in Intel Distribution for Python in Scikit-learn optimization.

Intel® oneAPI Data Analytics Library is an extension of Intel® Data Analytics Acceleration Library (Intel® DAAL).

Table of Contents

Build yours high-performance data science application with intel® oneDAL

Intel® oneDAL uses all capabilities of Intel® hardware, which allows you to get an sugnificant performance boost on the classic machine learning algorithms.

We provide highly optimized algorithmic building blocks for all stages of data analytics: preprocessing, transformation, analysis, modeling, validation, and decision making.

The current version of oneDAL provides Data Parallel C++ (DPC++) API extensions to the traditional C++ interface.

The size of the data is growing exponentially, as is the need for high-performance and scalable frameworks to analyze all this data and extract some benefits from it. Besides superior performance on a single node, the oneDAL distributed computation mode also provides excellent strong and weak scaling (check charts below).

Intel® oneDAL K-means fit, strong scaling result Intel® oneDAL K-means fit, weak scaling results

technical details: FPType: float32; HW: Intel Xeon Processor E5-2698 v3 @2.3GHz, 2 sockets, 16 cores per socket; SW: Intel® DAAL (2019.3), MPI4Py (3.0.0), Intel® Distribution Of Python (IDP) 3.6.8; Details available in the article https://arxiv.org/abs/1909.11822

Check out our examples and documentation for information about our API

Python API

Intel® oneDAL has a python API that is provided as a standalone python library called daal4py. Below is an example of how daal4py can be used for calculation KMeans clusters

import numpy as np
import pandas as pd
import daal4py as d4p

data = pd.read_csv("local_kmeans_data.csv", dtype = np.float32)

init_alg = d4p.kmeans_init(nClusters = 10,
                           fptype = "float",
                           method = "randomDense")

centroids = init_alg.compute(data).centroids
alg = d4p.kmeans(nClusters = 10, maxIterations = 50, fptype = "float",
                 accuracyThreshold = 0, assignFlag = False)
result = alg.compute(data, centroids)

Scikit-learn patching

Python interface to efficient Intel® oneDAL provided by daal4py allows one to create scikit-learn compatible estimators, transformers, clusterers, etc. powered by oneDAL which are nearly as efficient as native programs.

Speedups of Intel® oneDAL powered Scikit-learn over the original Scikit-learn, 28 cores, 1 thread/core
technical details: FPType: float32; HW: Intel(R) Xeon(R) Platinum 8276L CPU @ 2.20GHz, 2 sockets, 28 cores per socket; SW: scikit-learn 0.22.2, Intel® DAAL (2019.5), Intel® Distribution Of Python (IDP) 3.7.4; Details available in the article https://medium.com/intel-analytics-software/accelerate-your-scikit-learn-applications-a06cacf44912

daal4py have an API which matches API from scikit-learn. This framework allows you to speed up your existing projects by changing one line of code

from daal4py.sklearn.svm import SVC
from sklearn.datasets import load_digits

digits = load_digits()
X, y = digits.data, digits.target

svm = SVC(kernel='rbf', gamma='scale', C = 0.5).fit(X, y)
print(svm.score(X, y))

In addition daal4py provides an option to replace some scikit-learn methods by oneDAL solvers which makes it possible to get a performance gain without any code changes. This approach is the basis of Intel distribution for python scikit-learn. You can patch stock scikit-learn by using the only following commandline flag

python -m daal4py my_application.py

Patches can also be enabled programmatically:

from sklearn.svm import SVC
from sklearn.datasets import load_digits
from time import time

svm_sklearn = SVC(kernel="rbf", gamma="scale", C=0.5)

digits = load_digits()
X, y = digits.data, digits.target

start = time()
svm_sklearn = svm_sklearn.fit(X, y)
end = time()
print(end - start) # output: 0.141261...
print(svm_sklearn.score(X, y)) # output: 0.9905397885364496

from daal4py.sklearn import patch_sklearn
patch_sklearn() # <-- apply patch
from sklearn.svm import SVC

svm_d4p = SVC(kernel="rbf", gamma="scale", C=0.5)

start = time()
svm_d4p = svm_d4p.fit(X, y)
end = time()
print(end - start) # output: 0.032536...
print(svm_d4p.score(X, y)) # output: 0.9905397885364496

Distributed multi-node mode

Often data scientists require different tools for analysis regular and big data. daal4py offers various processing models, which makes it easy to enable distributed multi-node mode.

import numpy as np
import pandas as pd
import daal4py as d4p

d4p.daalinit() # <-- Initialize SPMD mode
data = pd.read_csv("local_kmeans_data.csv", dtype = np.float32)

init_alg = d4p.kmeans_init(nClusters = 10,
                           fptype = "float",
                           method = "randomDense",
                           distributed = True) # <-- change model to distributed

centroids = init_alg.compute(data).centroids

alg = d4p.kmeans(nClusters = 10, maxIterations = 50, fptype = "float",
                 accuracyThreshold = 0, assignFlag = False,
                 distributed = True)  # <-- change model to distributed

result = alg.compute(data, centroids)

For more details browse our daal4py documentation.

oneDAL Apache Spark MLlib samples

oneDAL provides scala / java interfaces that match Apache Spark MlLib API and use oneDAL solvers under the hood. This implementation allows you to get a 3-18X increase in performance compared to default Apache Spark MLlib.

technical details: FPType: double; HW: 7 x m5.2xlarge AWS instances; SW: Intel DAAL 2020 Gold, Apache Spark 2.4.4, emr-5.27.0; Spark config num executors 12, executor cores 8, executor memory 19GB, task cpus 8

Check samples tab for more details.

Installation

You can install oneDAL:

Installation from Source

See Installation from Sources for details.

Examples

Except C++ and Python API oneDAL also provide API for C++ SYCL and Java languages. Check out tabs below for more examples.

Documentation

Samples

Samples is an examples of how oneDAL can be used in different applications.

Technical Preview Features

Technical preview features are introduced to gain early feedback from developers. A technical preview feature is subject to change in the future releases. Using a technical preview feature in a production code base is therefore strongly discouraged. In C++ APIs, technical preview features are located in daal::preview and onedal::preview namespaces. In Java APIs, technical preview features are located in packages that have the com.intel.daal.preview name prefix. The preview features list:

  • Graph Analytics:
    • Undirected graph without edge and vertex weights (undirected_adjacency_array_graph) - 32bit vertex index only
    • Jaccard Similarity Coefficients for all vertex pairs, a batch algorithm which processes the graph by blocks

oneDAL and Intel® DAAL

Intel® oneAPI Data Analytics Library is an extension of Intel® Data Analytics Acceleration Library (Intel® DAAL).

This repository contains branches corresponding to both oneAPI and classical versions of the library. We encourage you to use oneDAL located under the master branch.

Product Latest release Branch Resources
oneDAL 2021.1-beta08 master
rls/onedal-beta08-rls
   Home page   
   Documentation   
   System Requirements
Intel® DAAL 2020 Gold rls/daal-2020-u2-rls    Home page   
   Developer Guide   
   System Requirements

About

Intel® Data Analytics Acceleration Library (Intel® DAAL)

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • C++ 60.4%
  • Java 18.5%
  • SWIG 15.3%
  • C 4.0%
  • Starlark 0.9%
  • Makefile 0.4%
  • Other 0.5%