Merge pull request #111 from daniel-habermann/Development

fix some typos and out-of-sync docstrings

bayesflow/computational_utilities.py

@@ -51,7 +51,7 @@ def posterior_calibration_error(
         The random draws from the approximate posteriors over ``num_datasets``
     prior_samples      : np.ndarray of shape (num_datasets, num_params)
         The corresponding ground-truth values sampled from the prior
-    alpha_resolution   : int, optional, default: 100
+    alpha_resolution   : int, optional, default: 20
         The number of credibility intervals (CIs) to consider
     aggregator_fun     : callable or None, optional, default: np.median
         The function used to aggregate the marginal calibration errors.
@@ -233,7 +233,7 @@ def mmd_kernel_unbiased(x, y, kernel):
     Returns
     -------
     loss   : tf.Tensor of shape (,)
-        The statistically unbiaserd squared maximum mean discrepancy (MMD) value.
+        The statistically unbiased squared maximum mean discrepancy (MMD) value.
     """
 
     m, n = x.shape[0], y.shape[0]

bayesflow/configuration.py

@@ -26,7 +26,7 @@
 
 
 class DefaultJointConfigurator:
-    """Fallback class for a generic configrator for joint posterior and likelihood approximation."""
+    """Fallback class for a generic configurator for joint posterior and likelihood approximation."""
 
     def __init__(self, default_float_type=np.float32):
         self.posterior_config = DefaultPosteriorConfigurator(default_float_type=default_float_type)

bayesflow/coupling_networks.py

@@ -313,7 +313,7 @@ def _calculate_spline(self, target, spline_params, inverse=False):
         target         : tf.Tensor of shape (batch_size, ..., dim_2)
             The target partition of the input vector to transform.
         spline_params  : tuple(tf.Tensor,...)
-            A tuple with tensors corresponding to the learnbale spline features:
+            A tuple with tensors corresponding to the learnable spline features:
             (left_edge, bottom_edge, widths, heights, derivatives)
         inverse        : bool, optional, default: False
             Flag indicating whether to run the block forward or backward.
@@ -516,7 +516,7 @@ def __init__(
             for the required entries.
         coupling_design       : str or callable, optional, default: 'affine'
             The type of internal coupling network to use. Must be in ['affine', 'spline'].
-            In general, spline couplings run slower than affine couplings, but require fewers coupling
+            In general, spline couplings run slower than affine couplings, but requires fewer coupling
             layers. Spline couplings may work best with complex (e.g., multimodal) low-dimensional
             problems. The difference will become less and less pronounced as we move to higher dimensions.
         permutation           : str or None, optional, default: 'fixed'
@@ -581,15 +581,15 @@ def __init__(
             self.act_norm = None
 
     def call(self, target_or_z, condition, inverse=False, **kwargs):
-        """Performs one pass through a the affine coupling layer (either inverse or forward).
+        """Performs one pass through the affine coupling layer (either inverse or forward).
 
         Parameters
         ----------
         target_or_z      : tf.Tensor
-            The estimation quantites of interest or latent representations z ~ p(z), shape (batch_size, ...)
+            The estimation quantities of interest or latent representations z ~ p(z), shape (batch_size, ...)
         condition        : tf.Tensor or None
             The conditioning data of interest, for instance, x = summary_fun(x), shape (batch_size, ...).
-            If `condition is None`, then the layer recuces to an unconditional ACL.
+            If `condition is None`, then the layer reduces to an unconditional ACL.
         inverse          : bool, optional, default: False
             Flag indicating whether to run the block forward or backward.
 

bayesflow/diagnostics.py

@@ -84,9 +84,9 @@ def plot_recovery(
         The parameter names for nice plot titles. Inferred if None
     fig_size          : tuple or None, optional, default : None
         The figure size passed to the matplotlib constructor. Inferred if None.
-    label_fontsize    : int, optional, default: 14
+    label_fontsize    : int, optional, default: 16
         The font size of the y-label text
-    title_fontsize    : int, optional, default: 16
+    title_fontsize    : int, optional, default: 18
         The font size of the title text
     metric_fontsize   : int, optional, default: 16
         The font size of the goodness-of-fit metric (if provided)
@@ -114,7 +114,7 @@ def plot_recovery(
     Raises
     ------
     ShapeError
-        If there is a deviation form the expected shapes of ``post_samples`` and ``prior_samples``.
+        If there is a deviation from the expected shapes of ``post_samples`` and ``prior_samples``.
     """
 
     # Sanity check
@@ -252,12 +252,12 @@ def plot_z_score_contraction(
 
     post_contraction = 1 - (posterior_variance / prior_variance)
 
-    In other words, the posterior is a proxy for the reduction in uncertainty gained by
+    In other words, the posterior contraction is a proxy for the reduction in uncertainty gained by
     replacing the prior with the posterior. The ideal posterior contraction tends to 1.
     Contraction near zero indicates that the posterior variance is almost identical to
     the prior variance for the particular marginal parameter distribution.
 
-    Note: Means and variances will be estimated vie their sample-based estimators.
+    Note: Means and variances will be estimated via their sample-based estimators.
 
     [1] Schad, D. J., Betancourt, M., & Vasishth, S. (2021).
     Toward a principled Bayesian workflow in cognitive science.
@@ -275,9 +275,9 @@ def plot_z_score_contraction(
         The parameter names for nice plot titles. Inferred if None
     fig_size          : tuple or None, optional, default : None
         The figure size passed to the matplotlib constructor. Inferred if None.
-    label_fontsize    : int, optional, default: 14
+    label_fontsize    : int, optional, default: 16
         The font size of the y-label text
-    title_fontsize    : int, optional, default: 16
+    title_fontsize    : int, optional, default: 18
         The font size of the title text
     tick_fontsize     : int, optional, default: 12
         The font size of the axis ticklabels
@@ -295,7 +295,7 @@ def plot_z_score_contraction(
     Raises
     ------
     ShapeError
-        If there is a deviation form the expected shapes of ``post_samples`` and ``prior_samples``.
+        If there is a deviation from the expected shapes of ``post_samples`` and ``prior_samples``.
     """
 
     # Sanity check for shape integrity
@@ -421,7 +421,7 @@ def plot_sbc_ecdf(
         The font size of the y-label and y-label texts
     legend_fontsize   : int, optional, default: 14
         The font size of the legend text
-    title_fontsize    : int, optional, default: 16
+    title_fontsize    : int, optional, default: 18
         The font size of the title text. Only relevant if `stacked=False`
     tick_fontsize     : int, optional, default: 12
         The font size of the axis ticklabels
@@ -587,11 +587,11 @@ def plot_sbc_histograms(
         The figure size passed to the matplotlib constructor. Inferred if None
     num_bins          : int, optional, default: 10
         The number of bins to use for each marginal histogram
-    binomial_interval : float in (0, 1), optional, default: 0.95
+    binomial_interval : float in (0, 1), optional, default: 0.99
         The width of the confidence interval for the binomial distribution
-    label_fontsize    : int, optional, default: 14
+    label_fontsize    : int, optional, default: 16
         The font size of the y-label text
-    title_fontsize    : int, optional, default: 16
+    title_fontsize    : int, optional, default: 18
         The font size of the title text
     tick_fontsize     : int, optional, default: 12
         The font size of the axis ticklabels
@@ -1077,7 +1077,7 @@ def plot_calibration_curves(
         The font size of the y-label and y-label texts
     legend_fontsize   : int, optional, default: 14
         The font size of the legend text (ECE value)
-    title_fontsize    : int, optional, default: 16
+    title_fontsize    : int, optional, default: 18
         The font size of the title text. Only relevant if `stacked=False`
     tick_fontsize     : int, optional, default: 12
         The font size of the axis ticklabels
@@ -1296,20 +1296,20 @@ def plot_mmd_hypothesis_test(
         The samples from the MMD sampling distribution under the null hypothesis "the model is well-specified"
     mmd_observed   : float
         The observed MMD value
-    alpha_level    : float
+    alpha_level    : float, optional, default: 0.05
         The rejection probability (type I error)
-    null_color     : str or tuple
+    null_color     : str or tuple, optional, default: (0.16407, 0.020171, 0.577478)
         The color of the H0 sampling distribution
-    observed_color : str or tuple
+    observed_color : str or tuple, optional, default: "red"
         The color of the observed MMD
-    alpha_color    : str or tuple
+    alpha_color    : str or tuple, optional, default: "orange"
         The color of the rejection area
-    truncate_vlines_at_kde: bool
+    truncate_vlines_at_kde: bool, optional, default: False
         true: cut off the vlines at the kde
         false: continue kde lines across the plot
-    xmin           : float
+    xmin           : float, optional, default: None
         The lower x-axis limit
-    xmax           : float
+    xmax           : float, optional, default: None
         The upper x-axis limit
     bw_factor      : float, optional, default: 1.5
         bandwidth (aka. smoothing parameter) of the kernel density estimate

bayesflow/helper_classes.py

@@ -162,7 +162,7 @@ class EarlyStopper:
     def __init__(self, patience=5, tolerance=0.05):
         """
 
-        patience          : int, optional, default: 4
+        patience          : int, optional, default: 5
             How many successive times the tolerance value is reached before triggering
             an early stopping recommendation.
         tolerance         : float, optional, default: 0.05
@@ -769,7 +769,7 @@ def __init__(self, capacity_in_batches=500):
 
         Parameters
         ----------
-        capacity_in_batches : int, optional, default: 50
+        capacity_in_batches : int, optional, default: 500
             The capacity of the buffer in batches of simulations. Could potentially grow
             very large, so make sure you pick a reasonable number!
         """

bayesflow/helper_functions.py

@@ -165,7 +165,7 @@ def backprop_step(input_dict, amortizer, optimizer, **kwargs):
     Parameters
     ----------
     input_dict  : dict
-        The configured output of the genrative model
+        The configured output of the generative model
     amortizer   : tf.keras.Model
         The custom amortizer. Needs to implement a compute_loss method.
     optimizer   : tf.keras.optimizers.Optimizer

bayesflow/helper_networks.py

@@ -179,7 +179,7 @@ def _inverse(self, target):
 
 
 class Orthogonal(tf.keras.Model):
-    """Imeplements a learnable orthogonal transformation according to [1]. Can be
+    """Implements a learnable orthogonal transformation according to [1]. Can be
     used as an alternative to a fixed ``Permutation`` layer.
 
     [1] Kingma, D. P., & Dhariwal, P. (2018). Glow: Generative flow with invertible 1x1
@@ -357,7 +357,7 @@ def call(self, target, inverse=False):
             If inverse=False: The transformed input and the corresponding Jacobian of the transformation,
             v shape: (batch_size, inp_dim), log_det_J shape: (,)
         target          :  tf.Tensor
-            If inverse=True: The inversly transformed targets, shape == target.shape
+            If inverse=True: The inversely transformed targets, shape == target.shape
 
         Notes
         -----

bayesflow/inference_networks.py

@@ -119,8 +119,8 @@ def __init__(
             Optional data-dependent initialization for the internal ``ActNorm`` layers, as done in [5]. Could be helpful
             for deep invertible networks.
         use_soft_flow         : bool, optional, default: False
-            Whether to perturb the taregt distribution (i.e., parameters) with small amount of independent
-            noise, as done in [2]. Could be helpful for degenrate distributions.
+            Whether to perturb the target distribution (i.e., parameters) with small amount of independent
+            noise, as done in [2]. Could be helpful for degenerate distributions.
         soft_flow_bounds      : tuple(float, float), optional, default: (1e-3, 5e-2)
             The bounds of the continuous uniform distribution from which the noise scale would be sampled
             at each iteration. Only relevant when ``use_soft_flow=True``.
@@ -178,7 +178,7 @@ def call(self, targets, condition, inverse=False, **kwargs):
         return self.forward(targets, condition, **kwargs)
 
     def forward(self, targets, condition, **kwargs):
-        """Performs a forward pass though the chain."""
+        """Performs a forward pass through the chain."""
 
         # Add noise to target if using SoftFlow, use explicitly
         # not in call(), since methods are public

bayesflow/sensitivity.py

@@ -53,11 +53,11 @@ def misspecification_experiment(
     second_config_dict  : dict
         Configuration for the second misspecification factor
         fields: name (str), values (1D np.ndarray)
-    error_function      : callable, default: bayesflow.computational_utilities.aggregated_rmse
+    error_function      : callable, default: bayesflow.computational_utilities.aggregated_error
         A callable that computes an error metric on the approximate posterior samples
     n_posterior_samples : int, optional, default: 500
         Number of samples from the approximate posterior per data set
-    n_sim               : int, optional, default: 100
+    n_sim               : int, optional, default: 200
         Number of simulated data sets per configuration
     configurator        : callable or None, optional, default: None
         An optional configurator for the misspecified simulations.

bayesflow/simulation.py

@@ -405,7 +405,7 @@ def __init__(
 
         Examples
         --------
-        Varying number of local factors (e.g., groups, participants) between 1 and 100::
+        Varying number of local factors (e.g., groups, participants) between 1 and 100:
 
             def draw_hyper():
                 # Draw location for 2D conditional prior
@@ -529,7 +529,7 @@ def __init__(self, batch_simulator_fun=None, simulator_fun=None, context_generat
         vectors and context variables and will pass the latter directly to the function. Power users should attempt to provide
         optimized batched simulators.
 
-        If a ``simulator_fun`` is provided, the interface will assume thatthe function operates on single parameter vectors and
+        If a ``simulator_fun`` is provided, the interface will assume that the function operates on single parameter vectors and
         context variables and will wrap the simulator internally to allow batched functionality.
 
         Parameters
@@ -815,9 +815,9 @@ def plot_pushforward(
         funcs_labels        : list of str
             A list of labels for the functions in funcs_list.
             The default behavior without user input is to call the functions "Aggregator function 1, Aggregator function 2, etc."
-        batch_size          : int
+        batch_size          : int, optional, default: 1000
             The number of prior draws to generate (and then create and visualizes simulations from)
-        show_raw_sims       : bool
+        show_raw_sims       : bool, optional, default: True
             Flag determining whether or not a plot of 49 raw (i.e. unaggregated) simulations is generated.
             Useful for very general data exploration.