feat: modified global options; added 1d optimization tests (#56)

* feat: modified global options; added 1d optimization tests

* fix: dev environment dependencies install.

---------

Co-authored-by: GurjeetSinghSangra <gurjeet.singh.sangra@gmail.com>

Author: lacerbi

environment.yml

@@ -1,7 +1,8 @@
 # run: conda env create --file environment.yml
 name: pybads-dev
+
 dependencies:
-  - python>=3.8
+  - python>=3.9
   - pip
   - pip:
-    - -r requirements.txt
+    - -e ".[dev]"

examples/pybads_example_1_basic_usage.ipynb

@@ -59,7 +59,7 @@
     "Formally, the goal of BADS is to *minimize* a target (or objective) function $f(\\mathbf{x}): \\mathbb{R}^D \\rightarrow \\mathbb{R}$, for $\\mathbf{x} \\in \\mathbb{R}^D$,\n",
     "\n",
     "$$\n",
-    "\\mathbf{x}^\\star = \\arg\\min_\\mathbf{x} f(\\mathbf{x}) \\qquad \\text{with} \\; \\text{lower_bounds}_d \\le x_d \\le \\text{upper_bounds}_d \\; \\text{ for } 1\\le d \\le D,\n",
+    "\\mathbf{x}^\\star = \\arg\\min_\\mathbf{x} f(\\mathbf{x}) \\qquad \\text{with} \\quad \\text{lower\\_bounds}_d \\le x_d \\le \\text{upper\\_bounds}_d \\; \\text{ for } 1\\le d \\le D,\n",
     "$$\n",
     "where $D$ is the dimensionality of the problem and `lower_bounds`, `upper_bounds` are arrays representing lower/upper bound constraints, which can be set to infinite for unbounded parameters."
    ]
@@ -131,18 +131,23 @@
       "     0           6               1               1         Initial mesh            Initial points\n",
       "     0          10               1             0.5         Refine grid             Train\n",
       "     1          18               1            0.25         Refine grid             Train\n",
-      "     2          22        0.250428            0.25     Successful search (ES-wcm)        \n",
-      "     2          24       0.0851394            0.25     Successful search (ES-ell)        \n",
-      "     2          26      0.00997961            0.25     Incremental search (ES-ell)        \n",
-      "     2          27     0.000224117            0.25     Incremental search (ES-ell)        \n",
-      "     2          34     0.000224117           0.125         Refine grid             \n",
-      "     3          42     0.000224117          0.0625         Refine grid             Train\n",
-      "     4          44     7.97867e-05          0.0625     Incremental search (ES-wcm)        \n",
-      "     4          50     7.97867e-05         0.03125         Refine grid             \n",
-      "     5          58     7.97867e-05       0.0078125         Refine grid             \n",
-      "     6          66     7.97867e-05      0.00195312         Refine grid             Train\n",
+      "     2          20        0.608156            0.25     Successful search (ES-ell)        \n",
+      "     2          21        0.408646            0.25     Successful search (ES-ell)        \n",
+      "     2          23        0.303906            0.25     Incremental search (ES-wcm)        \n",
+      "     2          25        0.212582            0.25     Incremental search (ES-ell)        \n",
+      "     2          30        0.212582           0.125         Refine grid             Train\n",
+      "     3          32        0.127714           0.125     Successful search (ES-wcm)        \n",
+      "     3          33       0.0139198           0.125     Successful search (ES-wcm)        \n",
+      "     3          34     0.000681342           0.125     Incremental search (ES-wcm)        \n",
+      "     3          42     0.000681342          0.0625         Refine grid             Train\n",
+      "     4          44     0.000120297          0.0625     Incremental search (ES-ell)        \n",
+      "     4          50     0.000120297         0.03125         Refine grid             Train\n",
+      "     5          58     0.000120297        0.015625         Refine grid             Train\n",
+      "     6          66     0.000120297      0.00390625         Refine grid             Train\n",
+      "     7          69     2.70979e-05      0.00390625     Incremental search (ES-wcm)        \n",
+      "     7          74     2.70979e-05     0.000976562         Refine grid             Train\n",
       "Optimization terminated: change in the function value less than options['tol_fun'].\n",
-      "Function value at minimum: 7.97867122044391e-05\n",
+      "Function value at minimum: 2.7097907605695835e-05\n",
       "\n"
      ]
     }
@@ -173,8 +178,8 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "BADS minimum at: x_min = [1.00744247 1.01543427], fval = 7.979e-05\n",
-      "total f-count: 67, time: 1.34 s\n"
+      "BADS minimum at: x_min = [0.99709786 0.993772  ], fval = 2.71e-05\n",
+      "total f-count: 75, time: 1.13 s\n"
      ]
     }
    ],
@@ -216,7 +221,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.9.16"
+   "version": "3.13.1"
   },
   "vscode": {
    "interpreter": {

pybads/bads/options.py

@@ -108,7 +108,8 @@ def load_options_file(
 
         # evaluation_parameters
         for key, val in evaluation_parameters.items():
-            exec(key + "=val")
+            global_scope = globals()
+            exec(f"{key} = {val}", global_scope)
 
         options_list = _read_config_file(options_path)
         for (key, value, description) in options_list:

pybads/testing/bads/test_bads_optimization.py

@@ -62,9 +62,37 @@ def test_univariate_input_and_opt():
     opt = BADS(rfn, x0, plausible_lower_bounds=plb, plausible_upper_bounds=pub)
     opt.optimize()
 
-def test_1D_opt():
-    D, x0, LB, UB, PLB, PUB, tol_errs = get_test_opt_conf(D=1)
-    fun = lambda x: np.sum((np.atleast_2d(x) / np.arange(1, len(x) + 1) ** 2) ** 2)
+def test_1D_opt_ndarray():
+    """Test optimization with 1D inputs and nd arrays"""
+    fun = lambda x: np.atleast_2d(x)**2
+    x0 = np.array([[2.0]])
+    LB = np.array([[-10.0]])
+    UB = np.array([[10.0]])
+    PLB = np.array([[-5.0]])
+    PUB = np.array([[5.0]])
+    tol_errs = np.array([0.1])
+    run_bads(fun, x0, LB, UB, PLB, PUB, tol_errs, f_min=0.0, assert_flag=True)
+
+def test_1D_opt_1darray():
+    """Test optimization with 1D inputs and 1D arrays"""
+    fun = lambda x: np.atleast_2d(x)**2
+    x0 = np.array([2.0])
+    LB = np.array([-10.0])
+    UB = np.array([10.0])
+    PLB = np.array([-5.0])
+    PUB = np.array([5.0])
+    tol_errs = np.array([0.1])
+    run_bads(fun, x0, LB, UB, PLB, PUB, tol_errs, f_min=0.0, assert_flag=True)
+
+def test_1D_opt_scalar():
+    """Test optimization with scalar inputs"""
+    fun = lambda x: np.atleast_2d(x)**2
+    x0 = 2.0
+    LB = -10.0
+    UB = 10.0
+    PLB = -5.0
+    PUB = 5.0
+    tol_errs = np.array([0.1])
     run_bads(fun, x0, LB, UB, PLB, PUB, tol_errs, f_min=0.0, assert_flag=True)
 
 def test_high_dim_opt():

pybads/testing/decorators/test_handle_0D_1D_input_decorator.py

@@ -49,6 +49,16 @@ def bar(self, x):
 
     x = np.ones(10)
     assert np.ndim(Foo().bar(x)) == 0
+    
+    # Test single value array
+    x = np.array([1.0])
+    assert np.ndim(Foo().bar(x)) == 0
+    assert isinstance(Foo().bar(x), float)
+    
+    # Test scalar input
+    x = 1.0
+    assert np.ndim(Foo().bar(x)) == 0
+    assert isinstance(Foo().bar(x), float)
 
 
 def test_1D_return_multiple_returns():

pybads/testing/variable_transformer/test_variable_transformer.py

@@ -136,6 +136,18 @@ def test_inverse_on_boundaries():
     X2 = np.ones((10, D)) * -10.
     assert np.all(np.isclose(X, X2))
 
+def test_1D_transform():
+    """Test 1D variable transformation"""
+    parameter_transformer = VariableTransformer(
+        D=1,
+        lower_bounds=np.array([[-10]]),
+        upper_bounds=np.array([[10]]),
+    )
+    X = np.array([[3]])
+    Y = parameter_transformer(X)
+    Y2 = np.array([[0.3]])
+    assert np.all(np.isclose(Y, Y2, atol=1e-04))
+
 def test_inverse_min_space():
     parameter_transformer = VariableTransformer(
         D=D,