Remove non-synthetic datasets from CI run to avoid overloading of external data sources

Author: Alexsandruss

configs/sklearn_example.json

@@ -30,16 +30,6 @@
                         "n_informative": "[SPECIAL_VALUE]0.5"
                     },
                     "split_kwargs": { "test_size": 0.2 }
-                },
-                {
-                    "source": "fetch_openml",
-                    "id": 1430,
-                    "split_kwargs": {
-                        "train_size": 1000,
-                        "test_size": 1000,
-                        "shuffle": true,
-                        "random_state": 42
-                    }
                 }
             ]
         },

configs/testing/azure-pipelines-ci.json

@@ -33,10 +33,6 @@
         "datasets": {
             "data":
             [
-                {
-                    "source": "fetch_openml",
-                    "id": 1430
-                },
                 {
                     "source": "make_classification",
                     "generation_kwargs": {

test-configuration-linux.yml

@@ -38,7 +38,7 @@ steps:
   - script: |
       source /usr/share/miniconda/etc/profile.d/conda.sh
       conda activate bench-env
-      python -m sklbench --report -l DEBUG --report -p algorithm:library=sklearn,sklearnex algorithm:estimator=PCA,KMeans,ElasticNet,KNeighborsClassifier data:dataset=skin_segmentation data:split_kwargs:train_size=5000 data:split_kwargs:test_size=5000
+      python -m sklbench --report -l DEBUG --report -p algorithm:library=sklearn,sklearnex algorithm:estimator=PCA,KMeans,ElasticNet,KNeighborsClassifier data:source=make_classification data:generation_kwargs:n_samples=10000 data:split_kwargs:train_size=0.5 data:split_kwargs:test_size=0.5
     displayName: CLI arguments example run
   - script: |
       source /usr/share/miniconda/etc/profile.d/conda.sh

test-configuration-win.yml

@@ -37,7 +37,7 @@ steps:
     condition: eq(variables['PKG_MANAGER'], 'conda')
   - script: |
       call activate bench-env
-      python -m sklbench --report -l DEBUG --report -p algorithm:library=sklearn,sklearnex algorithm:estimator=PCA,KMeans,ElasticNet,KNeighborsClassifier data:dataset=skin_segmentation data:split_kwargs:train_size=5000 data:split_kwargs:test_size=5000
+      python -m sklbench --report -l DEBUG --report -p algorithm:library=sklearn,sklearnex algorithm:estimator=PCA,KMeans,ElasticNet,KNeighborsClassifier data:source=make_classification data:generation_kwargs:n_samples=10000 data:split_kwargs:train_size=0.5 data:split_kwargs:test_size=0.5
     displayName: CLI arguments example run
   - script: |
       call activate bench-env