Intel® oneAPI Data Analytics Library
Community support for building compute-intensive applications that run fast on Intel® architecture.
Announcements
Welcome to the Intel Community. If you get an answer you like, please mark it as an Accepted Solution to help others. Thank you!
204 Discussions

## How to run daal example in python independently and monitor inputs and outputs of some main functions?

Beginner
491 Views

Hi,

I was able to run all examples in python using the following cmd:

PS C:\IntelPython3\share\pydaal_examples\examples\python> C:\IntelPython3\python.exe .\run_examples.py

What i am really interested in doing is to run just one example below and monitor inputs and outputs of some main functions.

C:\IntelPython3\share\pydaal_examples\examples\python\source\em\em_gmm_dense_batch.py

Here is what I have tried:

PS C:\IntelPython3\share\pydaal_examples\examples\python\source\em> C:\IntelPython3\python.exe .\em_gmm_dense_batch.py
Traceback (most recent call last):
File ".\em_gmm_dense_batch.py", line 53, in <module>
from utils import printNumericTable
ModuleNotFoundError: No module named 'utils'

So if all examples can run without error, i do not think something is missing. What is the correct way to run that example?

Thanks a lot!

5 Replies
Employee
491 Views

Hi cn,

The examples are set up to run from the share\pydaal_examples\examples\python directory, so to run a single example, you can do

cd C:\IntelPython3\share\pydaal_examples\examples\python
C:\IntelPython3\python.exe source\em\em_gmm_dense_batch.py


If you really want to run them from a different directory, just adjust the paths to the data files in the example, and make sure your PYTHONPATH points to the utils module, which is in C:\IntelPython3\share\pydaal_examples\examples\python\sources.

Chris

Beginner
491 Views

Chris,

Thanks for the info!

I was able to do that using Powershell and even better I could run run_examples.py in pyCharm!

If I expand source and run \em\em_gmm_dense_batch.py, it still returned me error:

C:\IntelPython3\python.exe C:/IntelPython3/share/pydaal_examples/examples/python/source/em/em_gmm_dense_batch.py
Traceback (most recent call last):
File "C:/IntelPython3/share/pydaal_examples/examples/python/source/em/em_gmm_dense_batch.py", line 67, in <module>
DataSourceIface.doDictionaryFromContext
File "C:\IntelPython3\lib\site-packages\daal\data_management\__init__.py", line 4946, in __new__
return FileDataSource_CSVFeatureManagerFloat64(*args)
File "C:\IntelPython3\lib\site-packages\daal\data_management\__init__.py", line 3944, in __init__
this = _data_management.new_FileDataSource_CSVFeatureManagerFloat64(*args)
SystemError: Error on file open

This example was runable in powershell but I hope to step through the code in pyCharm such that input and output in each line can be monitored.

Do you have any suggestions? Thanks

Moderator
491 Views

here is standalone example of em_gmm_dense_batch.py example - attached.

and here is the output

python.exe em_gmm_dense_batch.py
Weights
0.500     0.500

Means
10.166    0.045     -7.211
0.194     0.019     0.376

Covariance
7.568     1.306     -0.063
1.306     0.979     0.636
-0.063    0.636     2.277

Covariance
0.860     -0.229    0.081
-0.229    2.220     -0.102
0.081     -0.102    2.573

Beginner
491 Views

Thanks for the zip file, Gennady

I tried to run it in pyCharm and I was hoping to see some real time variable values but I was not able to. For instance, when I stepped to the line:

algorithm.input.setValues(em_gmm.inputValues, resultInit)

I hovered over 'resultInit' and looked for initialized mean and weight matrix, but I was showing nothing.

The same thing happened when i tried to look for input data matrix, and calculated initialized mean and weight matrix. So i am wondering if there is a way to see those values at run time? Are they all designed to be hidden and calculated under the hood?

Thanks!

Beginner
491 Views

if your error is "Error on file open", you can change the line 40 to

DAAL_PREFIX = jp('..', '..','..','data')

then your code will run normally.