Start adding new logic for library #53

Flaminietta · 2024-10-08T16:43:34Z

ProMCDA is refactored to become a library initiated in a Python environment by a few input parameters. There is no need for a configuration file, and intermediary and final calculations are exposed to the user.

ProMCDA is also refactored to be modular and flexible.

The refactoring facilitates the use of an API to call ProMCDA as a service.

mcda/models/ProMCDA.py

…keys in config settings

…ts, without robustness

…nd scores from dict to dataframe, broken tests

Flaminietta · 2024-11-12T16:56:06Z

@kapil-agnihotri, as discussed today, you can start having a look at the progress of transforming ProMCDA into a library and leave me any comments and suggestions. @mspada, you can also start familiarising yourself with the transformation of ProMCDA.

This is not a working version yet. From the commit where I changed the output format of the normalised criteria values and scores to DataFrames (instead of dictionaries) to expose them directly to the user calling the "normalize" and "aggregate" methods, the tests fail and you cannot use the library in the notebook. I am currently working on this.

My plan for the next release is to (in no particular order)

fix the tests and be able to instantiate ProMCDA and use its main methods (i.e.,normalize, aggregate, run, get_results);
clean up redundant code;
simplify the init of ProMCDA.py for the user who only wants to run the "normalize" and "aggregate" methods by making some parameters optional;
give the option to normalize either with zero values or only with values greater than zero;
give the option of outputting DataFrames or dictionaries (given my recent experience, this may not be that easy);
make logging optional/clean where needed;
eliminate input dictionaries and implement classes instead;
simplify the configuration logic and terminology (see this branch).

In the final version, all tests should run again and we need to proof all the functionalities with test data.

kapil-agnihotri · 2024-11-13T14:35:23Z

mcda/configuration/enums.py

+    MINIMUM = 'minimum'
+
+
+class NormalizationNames4Sensitivity(Enum):


I would refactor this enum to SensitivityNormalization

kapil-agnihotri · 2024-11-13T14:36:15Z

mcda/configuration/enums.py

+    RANK = 'rank'
+
+
+class OutputColumnNames4Sensitivity(Enum):


I would refactor it to SensitivityOutputColumns

kapil-agnihotri · 2024-11-13T14:44:11Z

mcda/configuration/enums.py

+    """
+    Names of output columns in case of sensitivity analysis
+    """
+    WS_MINMAX_01 = 'ws-minmax_01'


Applies to all the enums from L47 to L59:
Why do they have hyphen- and underscore _ both in values?
I would recommend to use underscore _ for all the values as it would increase readability and also will be as per the standards.
Although they are just output column names, still it would be great to avoid hyphen.

kapil-agnihotri · 2024-11-13T15:05:25Z

mcda/configuration/configuration_validator.py

+    :rtype: bool
+    """
+
+    keys_of_dict_values = {


It would be great to simplify these keys based on ranking service logic and open API specification that we created in the past.
robustness
monte_carlo

Also rename keys based on Open API specification.

kapil-agnihotri · 2024-11-13T15:07:26Z

mcda/configuration/configuration_validator.py

+
+    keys_of_dict_values = {
+        'sensitivity': ['sensitivity_on', 'normalization', 'aggregation'],
+        'robustness': ['robustness_on', 'on_single_weights', 'on_all_weights', 'given_weights', 'on_indicators'],


Also, it would be preferable to add these keys in enum instead of defining them here otherwise there would be multiple occurrence of these spread in your code.

kapil-agnihotri · 2024-11-13T18:03:15Z

mcda/configuration/configuration_validator.py

+        Tuple[List[str], None, None, dict]]:
+    """
+    Manage polarities and weights based on the specified robustness settings, ensuring that the appropriate adjustments
+    and normalizations are applied before returning the necessary data structures.


Any doc on return types?

kapil-agnihotri · 2024-11-13T18:03:52Z

mcda/configuration/configuration_validator.py

+    rand_weight_per_indicator = {}
+
+    # Managing polarities
+    if is_robustness_indicators == 0:


Again no need for explicit comparison to 0 or 1

kapil-agnihotri · 2024-11-13T18:06:19Z

mcda/models/ProMCDA.py

+        self.aggregated_matrix = None
+        self.ranked_matrix = None
+
+    def validate_inputs(self) -> Tuple[int, int, list, Union[list, List[list], dict], dict]:


I would expect this method to only validate and not to extract values

kapil-agnihotri · 2024-11-13T18:07:26Z

mcda/models/ProMCDA.py

+    def run_mcda(self, is_robustness_indicators: int, is_robustness_weights: int,
+                 weights: Union[list, List[list], dict]):
+        """
+        Execute the full ProMCDA process, either with or without uncertainties on the indicators.


incomplete docs

kapil-agnihotri · 2024-11-13T18:08:52Z

mcda/models/mcda_without_robustness.py

+        if method is None or method == NormalizationFunctions.TARGET.value:
+            indicators_target_01 = norm.target(feature_range=(0, 1))
+            indicators_target_without_zero = norm.target(feature_range=(0.1, 1))
+            add_normalized_df(indicators_target_01, "target_01")


hard coded string you can easily use enums here

Flaminietta added the enhancement New feature or request label Oct 8, 2024

Flaminietta requested review from mspada, thorsten-reitz and kapil-agnihotri October 8, 2024 16:43

Flaminietta self-assigned this Oct 8, 2024

refactor:start adding new logic for library

eae5fec

Flaminietta force-pushed the refactor_in_to_library branch from cef2ab3 to eae5fec Compare October 8, 2024 17:17

kapil-agnihotri reviewed Oct 9, 2024

View reviewed changes

mcda/models/ProMCDA.py Outdated Show resolved Hide resolved

Flaminietta added 6 commits October 10, 2024 17:06

refactor: add unit test for ProMCDA.py

61bf9a4

refactor: solve some unit tests, circual imports

993a211

unit test debugging 1

c8180db

refactor: add basic uni tests and remove open socket warning

7742a1a

refactor: delete config.py and add a function to validate dictionary …

9d747ca

…keys in config settings

refactor: add normalization as a method, failing tests

88e31c5

Flaminietta force-pushed the refactor_in_to_library branch from d232c1d to 88e31c5 Compare October 22, 2024 19:49

Flaminietta added 5 commits October 24, 2024 15:41

refactor: add enums for enumerations

d9e03b1

refactor: simplify normalize and aggregate methods and draft unit tes…

5c0cd1d

…ts, without robustness

add notebook for testing library function, change normalized_scores a…

1378354

…nd scores from dict to dataframe, broken tests

stop tracking files in demo_in_notebook/output_files/mock_output/

c274e4b

update gitignore

5344494

fix test_aggregate_with_specific_normalization_and_aggregation_methods

d980127

kapil-agnihotri requested changes Nov 13, 2024

View reviewed changes

Flaminietta added 7 commits November 14, 2024 16:18

fix all test_promcda

3852c11

test in notebook basic methods normlize and aggregate

8ca4f1d

improve column names in normalize

d312433

change and simplify the __init__, add optional and default parameters

7deee2d

modify normalize method and test for mcda without robustness

f7efe84

add normalization with robustness

fbf90ee

modify test for norm with robustness

93441d2

Flaminietta added 10 commits November 27, 2024 15:41

add and test normalization with robustness on indicators

a064fc0

finalize normalization

a770cc1

add aggregation with specific method

d587fe3

add aggregation without a specific method

d95aaa9

add aggregation with robustness

20cacda

fix unit test for aggregation, all variations of it

39c9061

update notebook with end to end tests

b9272a0

fixing aggregation with robustness

1d71997

fix aggregation with robustness on indicators and weights

17596e6

remve hard coded column names and use enums

eed0195

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Start adding new logic for library #53

Start adding new logic for library #53

Flaminietta commented Oct 8, 2024

Flaminietta commented Nov 12, 2024 •

edited

Loading

kapil-agnihotri Nov 13, 2024

kapil-agnihotri Nov 13, 2024

kapil-agnihotri Nov 13, 2024

kapil-agnihotri Nov 13, 2024

kapil-agnihotri Nov 13, 2024

kapil-agnihotri Nov 13, 2024

kapil-agnihotri Nov 13, 2024

kapil-agnihotri Nov 13, 2024

kapil-agnihotri Nov 13, 2024

kapil-agnihotri Nov 13, 2024

		MINIMUM = 'minimum'


		class NormalizationNames4Sensitivity(Enum):

Start adding new logic for library #53

Are you sure you want to change the base?

Start adding new logic for library #53

Conversation

Flaminietta commented Oct 8, 2024

Flaminietta commented Nov 12, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Flaminietta commented Nov 12, 2024 •

edited

Loading