integrate-module 0.95.0__tar.gz → 0.97.0__tar.gz

{integrate_module-0.95.0 → integrate_module-0.97.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: integrate_module
-Version: 0.95.0
+Version: 0.97.0
 Summary: Localized probabilistic data integration
 Author-email: Thomas Mejer Hansen <tmeha@geo.au.dk>
 License: MIT
@@ -46,6 +46,9 @@ This repository contains the INTEGRATE Python module for localized probabilistic
 
 Assuming you already have Python 3.10+ installed:
 
+`libaarhusxyz` must be installed manually before installing `integrate_module`, because the version that supports Python >= 3.12 is only available on GitHub (not PyPI):
+
+    pip install git+https://github.com/cultpenguin/libaarhusxyz
     pip install integrate_module
 
 On Windows, this will also install the Python wrapper for GA-AEM (1D EM forward modeling - GPL v2 code): [ga-aem-forward-win](https://pypi.org/project/ga-aem-forward-win/)

{integrate_module-0.95.0 → integrate_module-0.97.0}/README.md

@@ -13,6 +13,9 @@ This repository contains the INTEGRATE Python module for localized probabilistic
 
 Assuming you already have Python 3.10+ installed:
 
+`libaarhusxyz` must be installed manually before installing `integrate_module`, because the version that supports Python >= 3.12 is only available on GitHub (not PyPI):
+
+    pip install git+https://github.com/cultpenguin/libaarhusxyz
     pip install integrate_module
 
 On Windows, this will also install the Python wrapper for GA-AEM (1D EM forward modeling - GPL v2 code): [ga-aem-forward-win](https://pypi.org/project/ga-aem-forward-win/)

{integrate_module-0.95.0 → integrate_module-0.97.0}/integrate/__init__.py

@@ -70,6 +70,7 @@ from integrate.integrate_io import copy_hdf5_file
 from integrate.integrate_io import hdf5_scan
 from integrate.integrate_io import get_case_data
 from integrate.integrate_io import save_data_gaussian
+from integrate.integrate_io import xyz_to_h5
 from integrate.integrate_io import save_data_multinomial
 from integrate.integrate_io import write_data_gaussian  # Deprecated - use save_data_gaussian
 from integrate.integrate_io import write_data_multinomial  # Deprecated - use save_data_multinomial
@@ -122,6 +123,7 @@ from integrate.integrate_query import query_plot
 from integrate.integrate_query import save_query
 from integrate.integrate_query import load_query
 from integrate.integrate_query import get_prior_model_info
+from integrate.integrate_query import query_from_text
 
 # Import gex module functions
 from integrate.gex import read_gex as read_gex2

{integrate_module-0.95.0 → integrate_module-0.97.0}/integrate/integrate.py

@@ -88,11 +88,9 @@ def use_parallel(**kwargs):
 
     Parameters
     ----------
-    **kwargs : dict
-        Additional keyword arguments including showInfo for verbosity control.
-        showInfo : int, optional
-            If greater than 0, prints information about the environment and
-            parallel processing status. Default is 0.
+    showInfo : int, optional
+        If greater than 0, prints information about the environment and
+        parallel processing status. Default is 0.
 
     Returns
     -------
@@ -220,14 +218,12 @@ def integrate_update_prior_attributes(f_prior_h5, **kwargs):
     ----------
     f_prior_h5 : str
         The path to the HDF5 file to process.
-    **kwargs : dict
-        Additional keyword arguments.
-        showInfo : int, optional
-            Level of verbosity for output (default is 0).
+    showInfo : int, optional
+        Level of verbosity for output (default is 0).
     """
-    
+
     showInfo = kwargs.get('showInfo', 0)
-    
+
     # Check that hdf5 files exists
     if not os.path.isfile(f_prior_h5):
         if showInfo>=2:
@@ -338,11 +334,13 @@ def integrate_posterior_stats(f_post_h5='POST.h5', ip_range=None, **kwargs):
 
     For each **continuous** model parameter ``/Mx``:
 
-    - ``/Mx/Mean``     [Np, Nm]  Arithmetic mean of posterior realizations.
-    - ``/Mx/LogMean``  [Np, Nm]  Geometric mean (exp of mean of log values).
-    - ``/Mx/Median``   [Np, Nm]  Median of posterior realizations.
-    - ``/Mx/Std``      [Np, Nm]  Standard deviation of log10(posterior).
-    - ``/Mx/KL``       [Np, Nm]  KL divergence in bits. Only written when
+    - ``/Mx/Mean``          [Np, Nm]  Arithmetic mean of posterior realizations.
+    - ``/Mx/LogMean``       [Np, Nm]  Geometric mean (exp of mean of log values).
+    - ``/Mx/Median``        [Np, Nm]  Median of posterior realizations.
+    - ``/Mx/Std``           [Np, Nm]  Standard deviation of log10(posterior).
+    - ``/Mx/HarmonicMean``  [Np, Nm]  Trimmed harmonic mean: conductivity samples
+      are trimmed 10% each tail, averaged, then inverted back to resistivity.
+    - ``/Mx/KL``            [Np, Nm]  KL divergence in bits. Only written when
       ``computeKL_continuous=True``.
 
     For each **discrete** model parameter ``/Mx``:
@@ -474,6 +472,7 @@ def integrate_posterior_stats(f_post_h5='POST.h5', ip_range=None, **kwargs):
                 M_mean = np.full((nsounding, nm), np.nan)
                 M_std = np.full((nsounding, nm), np.nan)
                 M_median = np.full((nsounding, nm), np.nan)
+                M_harmonicmean = np.full((nsounding, nm), np.nan)
                 if computeKL_continuous:
                     M_KL = np.full((nsounding, nm), np.nan)
 
@@ -510,6 +509,10 @@ def integrate_posterior_stats(f_post_h5='POST.h5', ip_range=None, **kwargs):
                         M_median[iid,:] = np.median(m_post, axis=0)
                         with np.errstate(invalid='ignore', divide='ignore'):
                             M_std[iid,:] = np.std(np.log10(np.maximum(m_post, 1e-10)), axis=0)
+                        _c = 1.0 / np.maximum(m_post, 1e-10)
+                        _k = int(np.floor(0.10 * _c.shape[0]))
+                        _cs = np.sort(_c, axis=0)
+                        M_harmonicmean[iid, :] = 1.0 / np.mean(_cs[_k:_c.shape[0]-_k, :], axis=0)
                         if computeKL_continuous:
                             m_post_log = np.log10(np.maximum(m_post, 1e-10))
                             for _i in range(nm):
@@ -560,16 +563,23 @@ def integrate_posterior_stats(f_post_h5='POST.h5', ip_range=None, **kwargs):
                         # Geometric Mean: exp(mean(log(x)))
                         M_logmean[current_iids, :] = np.exp(np.mean(log_cube, axis=1))
                         
-                        # Std of Log10: 
+                        # Std of Log10:
                         # Math identity: std(log10(x)) = std(ln(x) / ln(10)) = std(ln(x)) * (1/ln(10))
                         # We reuse 'log_cube' and multiply by constant (faster than re-calculating log10)
                         M_std[current_iids, :] = np.std(log_cube, axis=1) * INV_LOG_10
 
+                        # Harmonic mean (trimmed 10% each tail in conductivity space)
+                        _c = 1.0 / np.maximum(m_cube, 1e-10)
+                        _nr = _c.shape[1]
+                        _k = int(np.floor(0.10 * _nr))
+                        _cs = np.sort(_c, axis=1)
+                        M_harmonicmean[current_iids, :] = 1.0 / np.mean(_cs[:, _k:_nr-_k, :], axis=1)
+
 
 
 
                 # Create datasets
-                for stat in ['Mean', 'Median', 'Std','LogMean']:
+                for stat in ['Mean', 'Median', 'Std', 'LogMean', 'HarmonicMean']:
                     if stat not in f_post:
                         dset = '/%s/%s' % (name,stat)
                         if dset not in f_post:
@@ -581,6 +591,7 @@ def integrate_posterior_stats(f_post_h5='POST.h5', ip_range=None, **kwargs):
                 f_post['/%s/%s' % (name,'Mean')][:] = M_mean
                 f_post['/%s/%s' % (name,'Median')][:] = M_median
                 f_post['/%s/%s' % (name,'Std')][:] = M_std
+                f_post['/%s/%s' % (name,'HarmonicMean')][:] = M_harmonicmean
                 if computeKL_continuous:
                     dset = '/%s/KL' % name
                     if dset not in f_post:
@@ -706,7 +717,7 @@ def sample_from_posterior(is_, d_sim, f_data_h5='tTEM-Djursland.h5', N_use=10000
             Temperature.
         EV : float
             Expected value.
-        is_ : int
+        is_index : int
             Index of the posterior sample.
     """
     with h5py.File(f_data_h5, 'r') as f:
@@ -862,13 +873,11 @@ def forward_gaaem(C=np.array(()),
         Path to GEX file. Default is None.
     showtime : bool, optional
         Flag to display execution time. Default is False.
-    **kwargs : dict
-        Additional keyword arguments.
-        showInfo : int, optional
-            Level of verbosity for output.
-        doCompress : bool, optional
-            Flag to enable layer compression. Default is True.
-    
+    showInfo : int, optional
+        Level of verbosity for output.
+    doCompress : bool, optional
+        Flag to enable layer compression. Default is True.
+
     Returns
     -------
     numpy.ndarray
@@ -1164,7 +1173,7 @@ def get_process_handle_count():
     import os
     return psutil.Process(os.getpid()).num_handles()
 
-def prior_data_gaaem(f_prior_h5, file_gex=None, stmfiles=None, N=0, doMakePriorCopy=True, im=1, id=1, im_height=0, Nhank=280, Nfreq=12, is_log=False, parallel=True, **kwargs):
+def prior_data_gaaem(f_prior_h5, file_gex=None, stmfiles=None, N=0, doMakePriorCopy=True, im=1, id=1, im_height=0, Nhank=280, Nfreq=12, is_log=False, parallel=True, force_replace=False, **kwargs):
     """
     Generate prior data for the GA-AEM method.
 
@@ -1202,6 +1211,10 @@ def prior_data_gaaem(f_prior_h5, file_gex=None, stmfiles=None, N=0, doMakePriorC
         Ncpu : int, optional
             Number of CPUs to use for parallel processing. Default is 0, which
             uses all available CPUs. Only used when parallel=True.
+        force_replace : bool, optional
+            If True, delete an existing /D{id} dataset before writing.
+            If False (default), print a warning and return early if the
+            dataset already exists.
         showInfo : int, optional
             Level of verbosity for output (0=silent, 1=normal, 2=verbose).
 
@@ -1304,27 +1317,27 @@ def prior_data_gaaem(f_prior_h5, file_gex=None, stmfiles=None, N=0, doMakePriorC
     Dname = '/D%d' % id
 
 
-    f_prior = h5py.File(f_prior_data_h5, 'a')
+    with h5py.File(f_prior_data_h5, 'r') as f_prior_r:
+        if im_height>0:
+            if (showInfo>1):
+                print('Using M%d for height' % im_height)
+            tx_height = f_prior_r[Mheight][:]
 
-    if im_height>0:    
-        if (showInfo>1):
-            print('Using M%d for height' % im_height)
-        tx_height = f_prior[Mheight][:]
+        # Get thickness
+        if 'x' in f_prior_r[Mname].attrs:
+            z = f_prior_r[Mname].attrs['x']
+        else:
+            z = f_prior_r[Mname].attrs['z']
+        thickness = np.diff(z)
 
-    # Get thickness
-    if 'x' in f_prior[Mname].attrs:
-        z = f_prior[Mname].attrs['x']
-    else:
-        z = f_prior[Mname].attrs['z']
-    thickness = np.diff(z)
+        # Get conductivity
+        if Mname in f_prior_r.keys():
+            C = 1 / f_prior_r[Mname][:]
+        else:
+            print('Could not load %s from %s' % (Mname, f_prior_data_h5))
 
-    # Get conductivity
-    if Mname in f_prior.keys():
-        C = 1 / f_prior[Mname][:]
-    else:
-        print('Could not load %s from %s' % (Mname, f_prior_data_h5))
+        N = f_prior_r[Mname].shape[0]
 
-    N = f_prior[Mname].shape[0]
     t1 = time.time()
     if not parallel:
         if (showInfo>-1):
@@ -1430,16 +1443,21 @@ def prior_data_gaaem(f_prior_h5, file_gex=None, stmfiles=None, N=0, doMakePriorC
         print('prior_data_gaaem: Time=%5.1fs/%d soundings. %4.1fms/sounding, %3.1fit/s' % (t_elapsed, N, 1000*t_elapsed/N,N/t_elapsed))
     
     # Write D to f_prior['/D1']
-    f_prior[Dname] = D
-
-    # Add method, type, file_ex, and im as attributes to '/D1'
-    f_prior[Dname].attrs['method'] = method
-    f_prior[Dname].attrs['type'] = type
-    f_prior[Dname].attrs['im'] = im
-    f_prior[Dname].attrs['Nhank'] = Nhank
-    f_prior[Dname].attrs['Nfreq'] = Nfreq
+    with h5py.File(f_prior_data_h5, 'a') as f_prior:
+        if Dname in f_prior:
+            if force_replace:
+                del f_prior[Dname]
+            else:
+                print("Key '%s' already exists in %s. Use force_replace=True to overwrite." % (Dname, f_prior_data_h5))
+                return f_prior_data_h5
+        f_prior[Dname] = D
 
-    f_prior.close()
+        # Add method, type, file_ex, and im as attributes to '/D1'
+        f_prior[Dname].attrs['method'] = method
+        f_prior[Dname].attrs['type'] = type
+        f_prior[Dname].attrs['im'] = im
+        f_prior[Dname].attrs['Nhank'] = Nhank
+        f_prior[Dname].attrs['Nfreq'] = Nfreq
 
     integrate_update_prior_attributes(f_prior_data_h5)
     
@@ -1462,13 +1480,11 @@ def prior_data_identity(f_prior_h5, id=0, im=1, N=0, doMakePriorCopy=False, **kw
         Number of soundings to consider. Default is 0 (use all).
     doMakePriorCopy : bool, optional
         Flag indicating whether to make a copy of the prior file. Default is False.
-    **kwargs : dict
-        Additional keyword arguments.
-        showInfo : int, optional
-            Level of verbosity for output.
-        forceDeleteExisting : bool, optional
-            Flag to force deletion of existing data. Default is True.
-    
+    showInfo : int, optional
+        Level of verbosity for output.
+    forceDeleteExisting : bool, optional
+        Flag to force deletion of existing data. Default is True.
+
     Returns
     -------
     str
@@ -1476,7 +1492,7 @@ def prior_data_identity(f_prior_h5, id=0, im=1, N=0, doMakePriorCopy=False, **kw
     """
     import integrate as ig
     import time
-    
+
     type = 'idenity'
     method = '--'
     showInfo = kwargs.get('showInfo', 0)
@@ -1599,12 +1615,10 @@ def prior_model_layered(lay_dist='uniform', dz = 1, z_max = 90,
         values below this threshold (including zero or negative values from 'normal'
         distribution) will be clamped to this value. Ensures physically realistic
         positive resistivity values. Default is 0.001 Ohm·m.
-    **kwargs : dict
-        Additional keyword arguments.
-        f_prior_h5 : str, optional
-            Path to the prior model file in HDF5 format. Default is ''.
-        showInfo : int, optional
-            Level of verbosity for output.
+    f_prior_h5 : str, optional
+        Path to the prior model file in HDF5 format. Default is ''.
+    showInfo : int, optional
+        Level of verbosity for output.
 
     Returns
     -------
@@ -1826,12 +1840,10 @@ def prior_model_workbench_direct(N=100000, RHO_dist='log-uniform', z1=0, z_max=
         values below this threshold (including zero or negative values from 'normal'
         distribution) will be clamped to this value. Ensures physically realistic
         positive resistivity values. Default is 0.001 Ohm·m.
-    **kwargs : dict
-        Additional keyword arguments.
-        f_prior_h5 : str, optional
-            Path to the prior model file in HDF5 format. Default is ''.
-        showInfo : int, optional
-            Level of verbosity for output.
+    f_prior_h5 : str, optional
+        Path to the prior model file in HDF5 format. Default is ''.
+    showInfo : int, optional
+        Level of verbosity for output.
 
     Returns
     -------
@@ -1961,12 +1973,10 @@ def prior_model_workbench(N=100000, p=2, z1=0, z_max= 100, dz=1,
         values below this threshold (including zero or negative values from 'normal'
         distribution) will be clamped to this value. Ensures physically realistic
         positive resistivity values. Default is 0.001 Ohm·m.
-    **kwargs : dict
-        Additional keyword arguments.
-        f_prior_h5 : str, optional
-            Path to the prior model file in HDF5 format. Default is ''.
-        showInfo : int, optional
-            Level of verbosity for output.
+    f_prior_h5 : str, optional
+        Path to the prior model file in HDF5 format. Default is ''.
+    showInfo : int, optional
+        Level of verbosity for output.
 
     Returns
     -------
@@ -2215,9 +2225,8 @@ def posterior_cumulative_thickness(f_post_h5, im=2, icat=[0], usePrior=False, **
             i_use[i,:]=np.arange(nr)
         
 
-    f_prior = h5py.File(f_prior_h5,'r')
-    M_prior = f_prior[Mstr][:]
-    f_prior.close()
+    with h5py.File(f_prior_h5,'r') as f_prior:
+        M_prior = f_prior[Mstr][:]
     nz = M_prior.shape[1]
 
     thick_mean = np.zeros((ns))
@@ -2352,72 +2361,48 @@ def synthetic_case(case='Wedge', **kwargs):
     case : str, optional
         The type of synthetic case to generate. Options are 'Wedge' and '3Layer'.
         Default is 'Wedge'.
-    **kwargs : dict
-        Additional parameters for synthetic case generation.
-
-        Common Parameters
-        -----------------
-        showInfo : int, optional
-            If greater than 0, print information about the generated case. Default is 0.
-        rho_1 : list or array_like, optional
-            Resistivity values for layer 1 along the profile. If a single value [v],
-            resistivity is constant at v. If multiple values [v1, v2, ...], resistivity
-            varies from v1 (left) to v2 (middle) to vN (right) using interpolation.
-            Only used when rho_1, rho_2, and rho_3 are all provided.
-        rho_2 : list or array_like, optional
-            Resistivity values for layer 2 along the profile. Same format as rho_1.
-        rho_3 : list or array_like, optional
-            Resistivity values for layer 3 along the profile. Same format as rho_1.
-
-        Parameters for 'Wedge' case
-        ---------------------------
-        x_max : int, optional
-            Maximum x-dimension size. Default is 1000.
-        dx : float, optional
-            Step size in the x-dimension. Default is 1000./x_max.
-        z_max : int, optional
-            Maximum z-dimension size. Default is 90.
-        dz : float, optional
-            Step size in the z-dimension. Default is 1.
-        z1 : float, optional
-            Depth at which the wedge starts. Default is z_max/10.
-        rho : list, optional
-            Density values for different layers. Default is [100, 200, 120].
-            Overridden by rho_1, rho_2, rho_3 if all three are provided.
-        wedge_angle : float, optional
-            Angle of the wedge in degrees. Default is 1.
-
-        Parameters for '3Layer' case
-        ----------------------------
-        x_max : int, optional
-            Maximum x-dimension size. Default is 100.
-        x_range : float, optional
-            Range in the x-dimension for the cosine function. Default is x_max/4.
-        dx : float, optional
-            Step size in the x-dimension. Default is 1.
-        z_max : int, optional
-            Maximum z-dimension size. Default is 60.
-        dz : float, optional
-            Step size in the z-dimension. Default is 1.
-        z1 : float, optional
-            Depth at which the first layer ends. Default is z_max/3.
-        z_thick : float, optional
-            Thickness of the second layer. Default is z_max/2.
-        rho1_1 : float, optional
-            Density at the start of the first layer. Default is 120.
-            Overridden by rho_1 if provided.
-        rho1_2 : float, optional
-            Density at the end of the first layer. Default is 10.
-            Overridden by rho_1 if provided.
-        rho2_1 : float, optional
-            Density at the start of the second layer. Default is rho1_2.
-            Overridden by rho_2 if provided.
-        rho2_2 : float, optional
-            Density at the end of the second layer. Default is rho1_1.
-            Overridden by rho_2 if provided.
-        rho3 : float, optional
-            Density of the third layer. Default is 120.
-            Overridden by rho_3 if provided.
+    showInfo : int, optional
+        If greater than 0, print information about the generated case. Default is 0.
+    rho_1 : list or array_like, optional
+        Resistivity values for layer 1 along the profile. If a single value [v],
+        resistivity is constant at v. If multiple values [v1, v2, ...], resistivity
+        varies from v1 (left) to v2 (middle) to vN (right) using interpolation.
+        Only used when rho_1, rho_2, and rho_3 are all provided.
+    rho_2 : list or array_like, optional
+        Resistivity values for layer 2 along the profile. Same format as rho_1.
+    rho_3 : list or array_like, optional
+        Resistivity values for layer 3 along the profile. Same format as rho_1.
+    x_max : int, optional
+        Maximum x-dimension size. Default is 1000 for 'Wedge', 100 for '3Layer'.
+    dx : float, optional
+        Step size in the x-dimension.
+    z_max : int, optional
+        Maximum z-dimension size. Default is 90 for 'Wedge', 60 for '3Layer'.
+    dz : float, optional
+        Step size in the z-dimension. Default is 1.
+    z1 : float, optional
+        Depth at which the wedge starts ('Wedge': default z_max/10) or first layer
+        ends ('3Layer': default z_max/3).
+    rho : list, optional
+        Density values for different layers ('Wedge' case). Default is [100, 200, 120].
+        Overridden by rho_1, rho_2, rho_3 if all three are provided.
+    wedge_angle : float, optional
+        Angle of the wedge in degrees ('Wedge' case). Default is 1.
+    x_range : float, optional
+        Range in the x-dimension for the cosine function ('3Layer' case).
+        Default is x_max/4.
+    z_thick : float, optional
+        Thickness of the second layer ('3Layer' case). Default is z_max/2.
+    rho1_1 : float, optional
+        Density at the start of the first layer ('3Layer'). Default is 120.
+    rho1_2 : float, optional
+        Density at the end of the first layer ('3Layer'). Default is 10.
+    rho2_1 : float, optional
+        Density at the start of the second layer ('3Layer'). Default is rho1_2.
+    rho2_2 : float, optional
+        Density at the end of the second layer ('3Layer'). Default is rho1_1.
+    rho3 : float, optional
+        Density of the third layer ('3Layer'). Default is 120.
 
     Returns
     -------

{integrate_module-0.95.0 → integrate_module-0.97.0}/integrate/integrate_borehole.py

@@ -623,10 +623,12 @@ def Pobs_to_datagrid(P_obs, X, Y, f_data_h5, r_data=10, r_dis=100, doPlot=False)
     2. Converts distance weight to temperature: T = 1 / w_dis
     3. Caps maximum temperature at 100 (very weak influence)
     4. For each grid point:
+
        - If T < 100: include point (i_use=1) and apply temperature scaling
        - If T ≥ 100: exclude point (i_use=0) and set observations to NaN
 
     Temperature scaling reduces probability certainty with distance:
+
     - T = 1 (close to observation): Original probabilities preserved
     - T > 1 (far from observation): Probabilities become more uniform
     - T ≥ 100 (very far): Observations effectively ignored