(Optional) Unit 3. Prepare Code¶

You can skip this unit if 2.3 Cumulative Update 1 has been applied to your environment.

Download Code Files¶

You can download the code files used in this tutorial. Extract the package, you can find the following code files included:

Number

Code File

Description

1

check_hive_config.py

Checks if EnOS Hive is available to save the prediction results.

2

generate_model_list.py

Generates a simulation instance list. A ParallelFor operator would perform parallel training or prediction tasks for each item in the list.

3

generate_random_int.py

Generates a random number from 1-10. In this tutorial, the raw data is ready for feature engineering and model training when the random number generated is smaller than 5.

4

generate_variables.py

Generates paths used in operators.

5

get_latest_model_version.py

Gets the latest model version information for a specific model.

6

prepare_data.py

Gets the raw data and split the training set and prediction set based on the sample_split_ratio defined in the pipeline.

7

prepare_predict_data.py

Simulates the prediction data processing.

8

prepare_train_data.py

Simulates the training data pre-processing and processing.

9

train_model.py

Trains a model with the training data and a specified algorithm Model training file. It also generates Mlflow model files.

10

write_results.py

Saves the prediction results as csv files.

11

requirements1.txt

Python package requirement file.

12

requirements2.txt

Python package requirement file.

Upload the Code Files¶

Since PythonEx operators are used in this tutorial, you need to upload the code files to the internal storage as follows:

From Enterprise Analytics Platform > Machine Intelligence Studio > MI Lab, click New Instance and enter the following values:
- Instance Name: winddemoinstance
- Workspace Storage: wind-demo-instance
Click the winddemo instance name to enter the JupyterLab environment.
Click New Folder icon to create a new folder named as wind_demo and click Upload File icon to upload code files to the notebook folder.
Open a new terminal and run the command pip install eap-notebook to install the eap-notebook package.
Run the command eap-notebook push -p wind_demo to upload the code files.

Next Unit¶

Design a Pipeline

Number	Code File	Description
1	check_hive_config.py	Checks if EnOS Hive is available to save the prediction results.
2	generate_model_list.py	Generates a simulation instance list. A ParallelFor operator would perform parallel training or prediction tasks for each item in the list.
3	generate_random_int.py	Generates a random number from 1-10. In this tutorial, the raw data is ready for feature engineering and model training when the random number generated is smaller than 5.
4	generate_variables.py	Generates paths used in operators.
5	get_latest_model_version.py	Gets the latest model version information for a specific model.
6	prepare_data.py	Gets the raw data and split the training set and prediction set based on the sample_split_ratio defined in the pipeline.
7	prepare_predict_data.py	Simulates the prediction data processing.
8	prepare_train_data.py	Simulates the training data pre-processing and processing.
9	train_model.py	Trains a model with the training data and a specified algorithm Model training file. It also generates Mlflow model files.
10	write_results.py	Saves the prediction results as csv files.
11	requirements1.txt	Python package requirement file.
12	requirements2.txt	Python package requirement file.