Merge pull request wichmann-lab#179 from otizonaizit/reinstate-parame…

…ter-estimate

Reinstate `Result.parameter_estimate`

docs/basic-usage.md

@@ -112,7 +112,7 @@ and the confidence intervals:
 jupyter:
   outputs_hidden: False
 ---
-print(result.get_parameters_estimate())
+print(result.parameter_estimate)
 ```
 
 This returns a python dictionary containing the estimated parameters.
@@ -132,7 +132,7 @@ The parameters estimated by psignifit are:
 Then, to obtain the threhsold you run
 
 ```{code-cell} ipython3
-print(result.get_parameters_estimate()['threshold'])
+print(result.parameter_estimate['threshold'])
 ```
 
 

docs/examples/parameter_recovery_demo.md

@@ -142,15 +142,15 @@ res = psignifit.psignifit(data, **options)
 
 +++ {"nteract": {"transient": {"deleting": false}}}
 
-Lastly, we can ensure that the values in our `res.get_parameters_estimate()` dictionary are equal to the values that we used to simulate them
+Lastly, we can ensure that the values in our `res.parameter_estimate` dictionary are equal to the values that we used to simulate them
 
 ```{code-cell} ipython3
 ---
 nteract:
   transient:
     deleting: false
 ---
-assert np.isclose(res.get_parameters_estimate()['lambda'], lambda_, atol=1e-2)
+assert np.isclose(res.parameter_estimate['lambda'], lambda_, atol=1e-2)
 ```
 
 ```{code-cell} ipython3
@@ -159,7 +159,7 @@ nteract:
   transient:
     deleting: false
 ---
-assert np.isclose(res.get_parameters_estimate()['gamma'], gamma, atol=1e-2)
+assert np.isclose(res.parameter_estimate['gamma'], gamma, atol=1e-2)
 ```
 
 ```{code-cell} ipython3
@@ -168,9 +168,9 @@ nteract:
   transient:
     deleting: false
 ---
-assert np.isclose(res.get_parameters_estimate()['eta'], 0, atol=1e-2)
-assert np.isclose(res.get_parameters_estimate()['threshold'], threshold, atol=1e-2)
-assert np.isclose(res.get_parameters_estimate()['width'], width, atol=1e-2)
+assert np.isclose(res.parameter_estimate['eta'], 0, atol=1e-2)
+assert np.isclose(res.parameter_estimate['threshold'], threshold, atol=1e-2)
+assert np.isclose(res.parameter_estimate['width'], width, atol=1e-2)
 ```
 
 ```{code-cell} ipython3

docs/examples/priors.md

@@ -191,7 +191,7 @@ res = ps.psignifit(data, experiment_type='2AFC')
 First lets have a look at the results with the standard prior strength:
 
 ```{code-cell} ipython3
-print('Fit:', res.get_parameters_estimate())
+print('Fit:', res.parameter_estimate)
 print('confidence Intervals:', res.confidence_intervals)
 ```
 
@@ -211,8 +211,8 @@ First see that the only parameter whose fit changes by this is the
 beta-variance parameter eta (the 5th)
 
 ```{code-cell} ipython3
-print('Fit with beta prior = 1: ', res1.get_parameters_estimate())
-print('Fit with beta prior = 200: ', res200.get_parameters_estimate())
+print('Fit with beta prior = 1: ', res1.parameter_estimate)
+print('Fit with beta prior = 200: ', res200.parameter_estimate)
 ```
 
 Now we have a look at the confidence intervals

docs/examples/results-object.md

@@ -42,7 +42,7 @@ The most important result are the fitted parameters of the psychometric
 function. They can be found in a dictionary format.
 
 ```{code-cell} ipython3
-print(res.get_parameters_estimate())
+print(res.parameter_estimate)
 ```
 
 For each of these parameters, also the confidence interval is contained
@@ -77,7 +77,7 @@ otherdata[:, 0] = otherdata[:, 0] + 0.01
 other_res = ps.psignifit(otherdata, res.configuration)
 
 # the difference in threshold should return the introduced shift
-print(other_res.get_parameters_estimate()['threshold'] - res.get_parameters_estimate()['threshold'])
+print(other_res.parameter_estimate['threshold'] - res.parameter_estimate['threshold'])
 ```
 
 ## Saving to JSON
@@ -94,7 +94,7 @@ It can be loaded again to be used at a later time, for example for plotting
 
 ```{code-cell} ipython3
 loaded_res = ps.Result.load_json(file_name)
-print(loaded_res.get_parameters_estimate())
+print(loaded_res.parameter_estimate)
 ```
 
 ```{code-cell} ipython3

docs/how_to/How-to-Fix-Parameters.md

@@ -51,6 +51,6 @@ res = ps.psignifit(data, **options)
 We print the parameter estimates to check that lambda was indeed fixed
 
 ```{code-cell} ipython3
-print(res.get_parameters_estimate())
+print(res.parameter_estimate)
 ```
 

docs/how_to/How-to-Get-Thresholds-and-Slopes.md

@@ -70,7 +70,7 @@ res.threshold(0.5, unscaled=True)
 which should be 0.0046, which is exactly the definition of the threshold we use in the fitting.
 
 ```{code-cell} ipython3
-res.get_parameters_estimate()['threshold']
+res.parameter_estimate['threshold']
 ```
 
 The function also returns worst-case credible intervals for the

psignifit/_result.py

@@ -19,8 +19,8 @@ def default(self, obj):
 
 @dataclasses.dataclass
 class Result:
-    parameters_estimate_MAP: Dict[str, float]
-    parameters_estimate_mean: Dict[str, float]
+    parameter_estimate_MAP: Dict[str, float]
+    parameter_estimate_mean: Dict[str, float]
     configuration: Configuration
     confidence_intervals: Dict[str, List[Tuple[float, float]]]
     data: NDArray[float]
@@ -68,7 +68,7 @@ def load_json(cls, file: Union[TextIO, str, Path], **kwargs: Any):
         result_dict['data'] = np.asarray(result_dict['data'])
         return cls.from_dict(result_dict)
 
-    def get_parameters_estimate(self, estimate_type: Optional[EstimateType]=None):
+    def get_parameter_estimate(self, estimate_type: Optional[EstimateType]=None):
         """ Get the estimate of the parameters by type.
 
         Args:
@@ -81,14 +81,16 @@ def get_parameters_estimate(self, estimate_type: Optional[EstimateType]=None):
             estimate_type = self.configuration.estimate_type
 
         if estimate_type == 'MAP':
-            estimate = self.parameters_estimate_MAP
+            estimate = self.parameter_estimate_MAP
         elif estimate_type == 'mean':
-            estimate = self.parameters_estimate_mean
+            estimate = self.parameter_estimate_mean
         else:
             raise ValueError("`estimate_type` must be either 'MAP' or 'mean'")
 
         return estimate
 
+    parameter_estimate = property(get_parameter_estimate)
+
     def threshold(self, proportion_correct: np.ndarray, unscaled: bool = False, return_ci: bool = True,
                   estimate_type: Optional[EstimateType]=None) -> Union[np.ndarray, List[Tuple[np.ndarray, np.ndarray]]]:
         """ Threshold stimulus value and confidence interval for a different proportion correct cutoff.
@@ -116,7 +118,7 @@ def threshold(self, proportion_correct: np.ndarray, unscaled: bool = False, retu
         proportion_correct = np.asarray(proportion_correct)
         sigmoid = self.configuration.make_sigmoid()
 
-        estimate = self.get_parameters_estimate(estimate_type)
+        estimate = self.get_parameter_estimate(estimate_type)
         if unscaled:  # set asymptotes to 0 for everything.
             lambd, gamma = 0, 0
         else:
@@ -152,7 +154,7 @@ def slope(self, stimulus_level: np.ndarray, estimate_type: Optional[EstimateType
         Returns:
             Slopes of the psychometric function at the stimulus levels.
         """
-        stimulus_level, param = np.asarray(stimulus_level), self.get_parameters_estimate(estimate_type)
+        stimulus_level, param = np.asarray(stimulus_level), self.get_parameter_estimate(estimate_type)
         sigmoid = self.configuration.make_sigmoid()
         return sigmoid.slope(stimulus_level, param['threshold'], param['width'], param['gamma'], param['lambda'])
 
@@ -174,7 +176,7 @@ def slope_at_proportion_correct(self, proportion_correct: np.ndarray, unscaled:
         stimulus_levels = self.threshold(proportion_correct, unscaled, return_ci=False, estimate_type=estimate_type)
         return self.slope(stimulus_levels)
 
-    def standard_parameters_estimate(self, estimate_type: Optional[EstimateType]=None):
+    def standard_parameter_estimate(self, estimate_type: Optional[EstimateType]=None):
         """ Get the parameters of the psychometric function in the standard format.
 
         `psignifit` uses the same intuitive parametrization, threshold and width, for all
@@ -192,6 +194,6 @@ def standard_parameters_estimate(self, estimate_type: Optional[EstimateType]=Non
             Standard parameters (loc, scale) for the sigmoid subclass.
         """
         sigmoid = self.configuration.make_sigmoid()
-        estimate = self.get_parameters_estimate(estimate_type)
+        estimate = self.get_parameter_estimate(estimate_type)
         loc, scale = sigmoid.standard_parameters(estimate['threshold'], estimate['width'])
         return loc, scale

psignifit/demos/parameters_estimates.ipynb

@@ -147,7 +147,6 @@
    "cell_type": "code",
    "execution_count": 5,
    "metadata": {
-    "collapsed": false,
     "jupyter": {
      "outputs_hidden": false,
      "source_hidden": false
@@ -310,7 +309,7 @@
    ],
    "source": [
     "# This gives the default estimate, in this case the mean estimate since we set it in the options.\n",
-    "result.get_parameters_estimate()  "
+    "result.parameter_estimate  "
    ]
   },
   {
@@ -334,8 +333,8 @@
     }
    ],
    "source": [
-    "# We can overwrite the default setting by specifying the `estimate_type` parameter\n",
-    "result.get_parameters_estimate(estimate_type='MAP')  "
+    "# We can overwrite the default setting by specifying the `estimate_type` parameter in the method get_parameter_estimate\n",
+    "result.get_parameter_estimate(estimate_type='MAP')  "
    ]
   },
   {
@@ -469,7 +468,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.9.6"
+   "version": "3.12.8"
   },
   "nteract": {
    "version": "0.28.0"

psignifit/psignifit.py

@@ -115,8 +115,8 @@ def psignifit(data: np.ndarray, conf: Optional[Configuration] = None,
         debug_dict['priors'] = priors
         debug_dict['bounds'] = bounds
 
-    return Result(parameters_estimate_MAP=estimate_MAP_dict,
-                  parameters_estimate_mean=estimate_mean_dict,
+    return Result(parameter_estimate_MAP=estimate_MAP_dict,
+                  parameter_estimate_mean=estimate_mean_dict,
                   configuration=conf,
                   confidence_intervals=intervals_dict,
                   parameter_values=grid,

psignifit/psigniplot.py

@@ -28,7 +28,7 @@ def plot_psychometric_function(result: Result,  # noqa: C901, this function is t
     if ax is None:
         ax = plt.gca()
 
-    params = result.get_parameters_estimate(estimate_type=estimate_type)
+    params = result.get_parameter_estimate(estimate_type=estimate_type)
     data = np.asarray(result.data)
     config = result.configuration
 
@@ -106,7 +106,7 @@ def _plot_residuals(x_values: np.ndarray,
                     estimate_type: EstimateType = None):
     if ax is None:
         ax = plt.gca()
-    params = result.get_parameters_estimate(estimate_type=estimate_type)
+    params = result.get_parameter_estimate(estimate_type=estimate_type)
     data = result.data
     sigmoid = result.configuration.make_sigmoid()
 
@@ -245,7 +245,7 @@ def plot_marginal(result: Result,
         ci_x = np.r_[CI[0], x[(x >= CI[0]) & (x <= CI[1])], CI[1]]
         ax.fill_between(ci_x, np.zeros_like(ci_x), np.interp(ci_x, x, marginal), color=line_color, alpha=0.5)
 
-        estimate = result.get_parameters_estimate(estimate_type=estimate_type)
+        estimate = result.get_parameter_estimate(estimate_type=estimate_type)
         param_value = estimate[parameter]
         ax.plot([param_value] * 2, [0, np.interp(param_value, x, marginal)], color='#000000')
 
@@ -306,7 +306,7 @@ def plot_prior(result: Result,
 
     data = result.data
     bounds = result.debug['bounds']
-    estimate = result.get_parameters_estimate(estimate_type=estimate_type)
+    estimate = result.get_parameter_estimate(estimate_type=estimate_type)
     sigmoid = result.configuration.make_sigmoid()
 
     sigmoid_x = np.linspace(bounds['threshold'][0], bounds['threshold'][1], 1000)
@@ -369,8 +369,8 @@ def plot_2D_margin(result: Result,
     if result.debug=={}:
         raise ValueError("Expects priors and marginals saved. Try running psignifit(....., debug=True).")
 
-    parameters_keys = result.parameters_estimate_MAP.keys()
-    parameter_indices = {param: i for i, param in enumerate(sorted(parameters_keys))}
+    parameter_keys = result.parameter_estimate_MAP.keys()
+    parameter_indices = {param: i for i, param in enumerate(sorted(parameter_keys))}
     other_param_ix = tuple(i for param, i in parameter_indices.items()
                            if param != first_param and param != second_param)
     marginal_2d = np.sum(result.debug['posteriors'], axis=other_param_ix)

psignifit/tests/test_basic.py

@@ -29,7 +29,7 @@ def get_std_options():
 def test_fit_basic(data):
     options = get_std_options()
     res = psignifit(data, **options)
-    param = res.parameters_estimate_MAP
+    param = res.parameter_estimate_MAP
     assert isclose(param['threshold'], 0.0046, abs_tol=0.0001)
     assert isclose(param['width'], 0.0045, abs_tol=0.0001)
     assert isclose(param['lambda'], 0.01, abs_tol=0.0001)
@@ -80,7 +80,7 @@ def test_bias_analysis(data):
 def test_fixed_parameters(data):
     options = get_std_options()
     res = psignifit(data, **options)
-    estim_param = res.parameters_estimate_MAP
+    estim_param = res.parameter_estimate_MAP
     fixed_param = res.configuration.fixed_parameters
     all_param_values = res.parameter_values