# pai-feature

Part of **PAI**

# Platform for AI (PAI) Experiment Management Console Guide

## Operations Overview

| Operation | Console Entry | Prerequisites | Description |
|----------|---------------|---------------|-------------|
| Compute Correlation Matrix | Console > Machine Learning Designer > Pipelines > Add Component > Correlation Coefficient Matrix | A dataset with numerical feature columns, Access to Machine Learning Designer in PAI | Adds and configures the Correlation Coefficient Matrix component to compute pairwise correlations between features |

## Operation Steps

### Compute Correlation Matrix

**Navigation**: Console > Machine Learning Designer > Pipelines > Add Component > Correlation Coefficient Matrix

**Prerequisites**:
- A dataset with numerical feature columns
- Access to Machine Learning Designer in PAI

1. Add the Correlation Coefficient Matrix component to your pipeline  
   - Element: **Correlation Coefficient Matrix** (button) — located in the Component library panel  
   - Notes: After clicking, the component appears on the pipeline canvas

2. Open the component’s settings panel  
   - Element: **Fields Setting** (tab) — located in the Settings panel on the right side of the screen  

3. Select feature columns for correlation calculation  
   - Element: **All Selected by Default** (checkbox) — located under the Fields Setting tab  
   - Notes: By default, all feature columns are selected. You can uncheck specific columns if needed.

4. Switch to the resource tuning tab  
   - Element: **Tuning** (tab) — located in the Settings panel  

5. Set the number of CPU cores for computation  
   - Element: **Cores** (text_input) — located under the Tuning tab  
   - Notes: Must be set together with Memory Size.

6. Set memory size per core  
   - Element: **Memory Size** (text_input) — located under the Tuning tab  
   - Notes: Valid values: 1024–65536 MB. Must be set together with Cores.

| Parameter | Type | Required | Options/Values | Description |
|-----------|------|----------|----------------|-------------|
| All Selected by Default | checkbox | No | — | Determines whether all feature columns are included in the correlation calculation by default. |
| Cores | number_input | No | — | Number of CPU cores allocated for the computation. Must be set together with Memory Size. |
| Memory Size | number_input | No | 1024–65536 MB | Memory allocated per core, in MB. Must be set together with Cores. |

## FAQ

Q: Where can I find the Correlation Coefficient Matrix component in the PAI console?  
A: Navigate to Console > Machine Learning Designer > Pipelines, then open the Component library panel and locate "Correlation Coefficient Matrix" under Statistical Analysis components.

Q: What happens if I leave the Cores and Memory Size fields empty?  
A: The system will use default resource allocation values. However, for large datasets, explicitly setting these ensures sufficient resources and avoids job failures.

Q: Can I modify the selected feature columns after adding the component?  
A: Yes. You can reopen the component’s settings panel at any time before running the pipeline and adjust the column selection under the Fields Setting tab.

Q: Are non-numerical columns included in the correlation calculation?  
A: No. Only numerical feature columns are considered. Non-numerical columns are automatically excluded from the correlation matrix computation.

Q: Do I need special permissions to use this component?  
A: You need standard access to Machine Learning Designer in PAI. No additional permissions are required beyond those needed to create and run pipelines.

## Pricing & Billing

### Billing Model
Billed per request (each time the component runs).

### Price Reference
| Tier | Input Price | Output Price |
|------|-------------|--------------|
| standard | 0.0001 / | 0.0001 / |

### Free Tier
100 free computations per month.

### Billing Notes
- Billed per run; output table storage fees are charged separately.
- Quota limits: maximum 1000 columns per run, up to 10,000 rows per task.