Arbitrary Waveform

.github/workflows/linux.yml

@@ -39,37 +39,37 @@ jobs:
           - python-version: "3.11"
             name: minimal
             os: ubuntu
-            conda: "numba 'scipy=1.10' scooby libdlf"
+            conda: "numba 'numpy==2.0.0' scipy scooby libdlf"
             test: ""
           - python-version: "3.12"
             name: plain
             os: ubuntu
-            conda: "numba 'scipy>=1.10' scooby libdlf"
+            conda: "numba 'numpy>=2.0.0' scipy scooby libdlf"
             test: ""
           - python-version: "3.12"
             name: full
             os: ubuntu
-            conda: "numba 'scipy>=1.10' scooby libdlf matplotlib pytest-mpl"
+            conda: "numba 'numpy>=2.0.0' scipy scooby libdlf matplotlib pytest-mpl"
             test: "--mpl"
           - python-version: "3.13"
             name: plain
             os: ubuntu
-            conda: "numba 'scipy>=1.10' scooby libdlf"
+            conda: "numba 'numpy>=2.0.0' scipy scooby libdlf"
             test: ""
           - python-version: "3.13"
             name: full
             os: ubuntu
-            conda: "numba 'scipy>=1.10' scooby libdlf matplotlib pytest-mpl"
+            conda: "numba 'numpy>=2.0.0' scipy scooby libdlf matplotlib pytest-mpl"
             test: "--mpl"
           - python-version: "3.14"
             name: plain
             os: ubuntu
-            conda: "numba 'scipy>=1.10' scooby libdlf"
+            conda: "numba 'numpy>=2.0.0' scipy scooby libdlf"
             test: ""
           - python-version: "3.14"
             name: full
             os: ubuntu
-            conda: "numba 'scipy>=1.10' scooby libdlf matplotlib pytest-mpl"
+            conda: "numba 'numpy>=2.0.0' scipy scooby libdlf matplotlib pytest-mpl"
             test: "--mpl"
 
     env:

CHANGELOG.rst

@@ -6,23 +6,59 @@ Version 2
 ~~~~~~~~~
 
 
-v2.5.x
+v2.6.x
 """"""
 
 
-*latest*
---------
+v2.6.0: Arbitrary waveforms
+---------------------------
+
+**2026-01-31**
+
+This version finally introduces the possibility for arbitrary waveforms.
 
 - Modelling routines:
 
+  - ``bipole`` and ``dipole``:
+
+    - Arbitrary waveforms: Time-domain modelling can new be done for any
+      arbitrary piecewise linear waveform, in addition to impulse
+      (``signal=0``), step-on (``signal=1``), and step-off (``signal=-1``)
+      responses.
+    - User-defined bandpass filters can now be applied to the frequency domain
+      result through user-provided functions.
+
   - Merged ``loop`` into ``bipole``; there is no need for a special routine.
     Simply use ``bipole`` with ``msrc='b'`` and ``mrec=True`` for the same
-    effect. As a result, ``empymod.utils.check_bipole`` changed its signature.
-    For ``bipole``, it new prints the source and receiver type (electric field,
-    magnetic field, magnetic flux, or electric current). Both,
-    ``empymod.model.loop`` and ``empymod.utils.check_bipole``'s old signature
-    are deprecated and will be removed in v3.
+    effect.
+
+- New prints (if verbose):
+
+  - The source/receiver types are new printed.
+  - The signal/waveform is new printed.
+
+- New function ``empymod.utils.check_waveform``, to check the waveform.
 
+- New deprecations (will be removed in v3):
+
+  - ``empymod.model.loop`` will be removed in v3.
+  - ``empymod.utils.check_bipole``: New signature, it now returns also the
+    src/rec field and
+    type; old signature will be removed in v3.
+  - ``empymod.utils.check_time``: New signature, it now also returns the
+    signal; old signature will be removed in v3.
+  - ``empymod.utils.check_time_only``: New signature, it now also returns the
+    signal; old signature will be removed in v3.
+
+- Bumped the minimum requirements to:
+
+  - Python 3.11
+  - NumPy 2.0
+  - SciPy, Numba, libdlf, scooby (without minimum version)
+
+
+v2.5.x
+""""""
 
 
 v2.5.4: Bugfix ext. fcts with z+ up
@@ -143,7 +179,7 @@ The code is now compatible with NumPy v2.
 
 - Gallery Update Part I:
 
-  - Update for Jupyterlab (ipympl/widget) 
+  - Update for Jupyterlab (ipympl/widget)
   - Replaced implicit by explicit pyplots
   - Use by default a positiv z-upwards coordinate system
   - Part I: frequency domain; reproducing; published

CREDITS.rst

@@ -16,8 +16,8 @@ namely:
   Aleksander Mousatov.
 - 2018-2021: `Delft University of Technology <https://www.tudelft.nl>`_;
   through the *Gitaro.JIM* project (till 05/2021, emg3d v1.0.0), funded by
-  `MarTERA <https://www.martera.eu>`_ as part of Horizon 2020, a funding scheme
-  of the European Research Area.
+  MarTERA as part of Horizon 2020, a funding scheme of the European Research
+  Area.
 - 2021-2022: `TERRASYS Geophysics GmbH & Co. KG
   <https://www.terrasysgeo.com>`_.
 - 2021-2024: `Delft University of Technology <https://www.tudelft.nl>`_ through

docs/manual/info.rst

@@ -154,8 +154,8 @@ the example
 :ref:`sphx_glr_gallery_educational_dlf_standard_lagged_splined.py`.
 
 
-Looping
--------
+Looping to reduce memory
+------------------------
 
 By default, you can compute many offsets and many frequencies all in one go,
 vectorized (for the *DLF*), which is the default. The ``loop`` parameter gives
@@ -218,6 +218,24 @@ A common alternative to this trick is to apply a lowpass filter to filter out
 the unstable high frequencies.
 
 
+Loops (FEM/TEM)
+---------------
+
+The kernel of ``empymod`` computes the electric and magnetic Green's functions
+for infinitesimal small dipoles. Combining various of these small dipoles
+allows to model any kind of source and receiver configurations, including
+loops. To make the correct assembly of dipoles is not always trivial, and the
+easiest way is probably to look for a similar example and modify it. Here a
+list of loop examples in the gallery:
+
+- :ref:`sphx_glr_gallery_fdomain_ipandq.py` (GEM-2 & DUALEM-842);
+- :ref:`sphx_glr_gallery_tdomain_ip_vrm.py`;
+- :ref:`sphx_glr_gallery_tdomain_tem_temfast.py` (TEM-FAST);
+- :ref:`sphx_glr_gallery_tdomain_tem_walktem.py` (Walk-TEM);
+- :ref:`sphx_glr_gallery_educational_dipoles_and_loops.py`;
+- :ref:`sphx_glr_gallery_reproducing_ward1988.py`.
+
+
 Hook for user-defined computation of :math:`\eta` and :math:`\zeta`
 -------------------------------------------------------------------
 
@@ -266,10 +284,45 @@ characteristics:
 And then you call ``empymod`` with ``res={'res': res-array, 'tau': tau,
 'func_eta': my_new_eta}``.
 
-Have a look at the corresponding example in the Gallery, where this hook is
-exploited in the low-frequency range to use the Cole-Cole model for IP
-computation. It could also be used in the high-frequency range to model
-dielectricity.
+Have a look at, e.g., the :ref:`sphx_glr_gallery_tdomain_cole_cole_ip.py`
+example in the Gallery, where this hook is exploited in the low-frequency range
+to use the Cole-Cole model for IP computation. It could also be used in the
+high-frequency range to model dielectricity.
+
+
+Hook for user-defined bandpass filter
+-------------------------------------
+
+Similar to the hooks described above, there is since ``empymod v2.6.0`` a hook
+to apply a bandpass filter (or any function) to the frequency domain result, by
+proving a specific dictionary as ``bandpass`` argument. The signature of the
+function must be ``func(inp, p_dict)``, where ``inp`` is the dictionary you
+provide, and ``p_dict`` is a dictionary that contains all parameters so far
+computed in empymod ``[locals()]``. Any change to the frequency domain result
+must be done in-place, and the function does not return anything. Refer to the
+time-domain loop examples in the gallery. The dictionary must contain at least
+the keyword ``'func'``, containing the actual function, but can contain any
+other parameters too.
+
+**Dummy example**
+
+.. code-block:: python
+
+    def my_bandpass(inp, p_dict):
+        # Your computations, using the parameters you provided
+        # in `inp` and the parameters from empymod in `p_dict`.
+        # In the example line below, we provide, e.g.,  inp["cutoff"]
+        # Modification must happen in-place to "EM", nothing is returned.
+
+        # Here just a simple cutoff to show the principle.
+        p_dict["EM"][p_inp["freq"] > inp["cutoff"] *= 0
+
+And then you call ``empymod`` with ``bandpass={"cutoff": 1e6, "func":
+my_bandpass}``.
+
+Have a look at, e.g., the :ref:`sphx_glr_gallery_tdomain_tem_walktem.py`
+example in the Gallery, where this hook is applied.
+
 
 
 Zero horizontal offset

docs/manual/installation.rst

@@ -13,8 +13,8 @@ or via ``pip``:
 
    pip install empymod
 
-Requirements are the modules ``numpy``, ``scipy``, ``numba``, ``libdlf``, and
-``scooby``.
+Requirements are the modules ``numpy>=2.0.0``, ``scipy``, ``numba``,
+``libdlf``, and ``scooby``.
 
 The modeller empymod comes with add-ons (``empymod.scripts``). These add-ons
 provide some very specific, additional functionalities. Some of these add-ons

docs/manual/references.rst

@@ -20,7 +20,7 @@ References
 .. [Ghos70] Ghosh, D. P.,  1970, The application of linear filter theory to the
    direct interpretation of geoelectrical resistivity measurements: Ph.D.
    Thesis, TU Delft; UUID: `88a568bb-ebee-4d7b-92df-6639b42da2b2
-   <http://resolver.tudelft.nl/uuid:88a568bb-ebee-4d7b-92df-6639b42da2b2>`_.
+   <https://resolver.tudelft.nl/uuid:88a568bb-ebee-4d7b-92df-6639b42da2b2>`_.
 .. [GuSi97] Guptasarma, D., and B. Singh, 1997, New digital linear filters for
    Hankel J0 and J1 transforms: Geophysical Prospecting, 45, 745--762; DOI:
    `10.1046/j.1365-2478.1997.500292.x
@@ -42,9 +42,7 @@ References
 .. [Key09] Key, K., 2009, 1D inversion of multicomponent, multifrequency marine
    CSEM data: Methodology and synthetic studies for resolving thin resistive
    layers: Geophysics, 74(2), F9--F20; DOI: `10.1190/1.3058434
-   <https://doi.org/10.1190/1.3058434>`_. Software:
-   `marineemlab.ucsd.edu/Projects/Occam/1DCSEM
-   <https://marineemlab.ucsd.edu/Projects/Occam/1DCSEM>`_.
+   <https://doi.org/10.1190/1.3058434>`_.
 .. [Key12] Key, K., 2012, Is the fast Hankel transform faster than quadrature?:
    Geophysics, 77(3), F21--F30; DOI: `10.1190/geo2011-0237.1
    <https://doi.org/10.1190/geo2011-0237.1>`_; Software: