Overview

We translate EO files into initial PHI programs. Next, we normalize these programs and get normalized PHI programs. Then, we collect metrics for initial and normalized PHI programs.

Metrics

An EO file contains multiple test objects. After translation, these test objects become attributes in PHI programs. We call these attributes "tests".

We collect metrics on the number of Dataless formations, Applications, Formations, Dispatches in tests. We want normalized tests to have less such elements than initial tests do.

A metric change for a test is calculated by the formula

(metric_initial - metric_normalized) / metric_initial

where:

• metric_initial is the metric for the initial test
• metric_normalized is the metric for the normalized test

Expected

Metric changes are expected to be as follows or greater:

• Dataless formations: 20.00%
• Applications: 20.00%
• Formations: 10.00%
• Dispatches: 20.00%

We expect such changes for at least 90.00% of tests.

Actual

We normalized 456 tests.

All metrics were improved for 333 (73.03%) tests.

Tests where a particular metric was improved:

• Dataless formations: 383 (83.99%)
• Applications: 333 (73.03%)
• Formations: 383 (83.99%)
• Dispatches: 352 (77.19%)

Table

The table below provides detailed information about tests.

Columns

Columns in this table are sortable.

Hover over a header cell from the second row of header cells (Attribute Initial, etc.) to see a triangle demonstrating the sorting order.

• ▾: descending
• ▴: ascending
• ▸: unordered

Click on the triangle to change the sorting order in the corresponding column.