Merico DevInsight

https://merico.cn/
Admin
Merico DevInsight helps hundreds of clients quickly and cost-effectively build R&D performance...
Tokens:82,409
Snippets:1,327
Trust Score:6.2
Update:4 months ago
Show doc for...
Tokens:
Raw
### Code Equivalent vs. Lines of Code: Example 2 (Code Block Movement)

Source: https://merico.cn/blog/equivalent-line-of-code

Illustrates that LOC fails to detect code block movements, while ELOC accurately assigns zero value to such operations. This example involves swapping the order of functions within a class.

```text
Example 2

Code lines are not good at detecting code block movements. 
For example, the code change below, simply swapping the order of functions in the class will result in 4 lines added and 4 lines deleted.
However, from the perspective of the abstract syntax tree, this modification only changes the order of the node corresponding to the myMethod() function under its parent node, and the node itself has not been modified. Therefore, the code equivalent of modifying myMethod() is 0.
```

--------------------------------

### Code Equivalent vs. Lines of Code: Example 3 (Differentiating Code Effort)

Source: https://merico.cn/blog/equivalent-line-of-code

Highlights LOC's inability to distinguish the effort required for different types of code, whereas ELOC uses weighted AST nodes to better quantify development work. This Python example compares code for finding symmetric pairs with test data.

```python
Example 3

Consider the following Python code, its function is to find symmetric pairs in a given dictionary. The test data test_dict and the actual function find_sym_pairs() contribute an equal number of lines of code (7 lines), which of course cannot reflect the different workload of writing these two pieces of code.
However, by assigning different weights to each AST node type, we can more reasonably evaluate the editing operations of different types of AST nodes and quantify the workload in the development process more reasonably.
```

--------------------------------

### Code Review Activity Trend Analysis (Partial Code)

Source: https://merico.cn/blog/improvements-in-code-reviews

Calculates weekly creation counts, comment counts, and coverage rates. This snippet shows the setup for plotting a combined bar and line chart to visualize trends.

```Python
# Assuming 'create_data' and 'comment_data' are preprocessed and merged with '年月'
# This is a conceptual snippet, actual implementation requires detailed data aggregation
weekly_trends = create_data.groupby('年月').agg(创建数=('iid', 'count')) # Simplified example
# Further aggregation for comments and coverage would follow
# Then plotting using plt.bar and plt.plot
```
Tokens:
Raw