PROPOSOAL: different anonymization methods for qi

allow in `sigo` configuration to choose the anonymization method for an attribute :

```yaml
version: "2"

kAnonymity: 3
lDiversity: 2
sensitives:
- s
rules:
- name: fieldA
aggregation: mean
- name: fieldB
aggregation: randomNoise
- name: fieldC
aggregation: median
```

No source file or test is named. Start by locating how SIGO reads its YAML configuration and applies each rule's aggregation; use the example as the target shape. Done means an attribute can select an anonymization method such as mean, randomNoise, or median through configuration.

Written by the indexing model from the issue text.