hyvr.geo package

Submodules

hyvr.geo.ae_realization module

Base class for AE realizations

class hyvr.geo.ae_realization.AERealization

Bases: object

AE realizations are the realization of a certain AE type in a HyVR simulation. For example, in the ‘made.ini’ test case, the lowest stratum consists of ‘clay_sheet’ architectural elements. While ‘clay_sheet’ is an AE type, each of these sheets is an AE realization.

This is the base class for AE realizations. Any instantiated object should be an instance of one of the (currently) three subtypes: TroughAE, ChannelAE, and SheetAE.

Typically these objects should not be created directly, but via the generate_realization method of the respective AEType object.

An AE realization contains a list of geometrical objects that are typical for this AE. For example, a TroughAE contains a list of Trough objects. These objects should be implemented in separate files as cython cdef classes, similar to the Trough class.

AERealizations are designed to be used from Python and from Cython code. Therefore, after construction the object list is converted to multiple arrays of object properties.

For example, a TroughAE has a list of Trough objects, and each Trough object has a property a of type float (lenth of semi-major axis of ellipsoid). Therefore, the TroughAE object also has a float array called object_a as property. This makes access from Cyton code faster, because we can then iterate over float arrays, instead of iterating over Python lists.

The derived AERealization types must provide the methods generate_object, maybe_assign_points_to_objects and maybe_assign_points_to_object.

bg_azim
bg_dip
bg_facies
bottom_surface
create_object_arrays()

This method is called after the object list has been created, so after all objects have been generated.

Its main purpose is to create fixed size arrays of object attributes as members of the AERealization subtype, so that they can be accessed without needing to access the python list of objects.

All subclasses of AERealization must override this method.

generate_objects()

This is the method that should be used for generating geometrical objects of the right type. Override this in all of your subclasses.

It’s very important that the objects are stored in the order such that the object with the highest zmax is the first one, i.e. such that we can go from top to bottom by iterating over the list of objects.

To add the objects to the object list, use the _add_object function below

maybe_assign_points_to_object()

This function checks whether the current grid cell with given coordinates is inside the trough and assigns facies, azimuth and dip by altering the passed arrays.

Parameters

  • oi (int) – object index

  • geo_ids (np.int32 array of size 4) – This holds the geological indices, i.e. facies number, architectural element number, hydrofacies assemblage (ha) and hydrofacies assemblage type (hat).

  • angles (double array) – array of size 2 that holds the azimuth and dip of the cell. This will also be altered.

  • y, z (x,) – cell coordinates

  • y_idx (x_idx,) – indices of x and y position in grid.

  • grid (Grid object) –

n_objects
num
object_azim
object_dip
object_facies
object_facies_array
object_list
object_num_facies
object_num_ha
object_zmax_list
object_zmaxs
object_zmin_list
object_zmins
top_surface
type_id
type_name
type_params
zmax
zmin

hyvr.geo.ae_types module

This module contains the architectural element types.

class hyvr.geo.ae_types.AEType(ae_section, name)[source]

Bases: object

Architectural element types are the kind of things that are read from the supplied ini-file. For example, in the made.ini testcase, there are the architectural element types “clay_sheet”, “sand_sheet”, “clay_lens”, “crossbedded_scour”, … In HyVR, these are instances of the AEType class.

The actual AEs that are realized in a HyVR simulation are instances of AERealization and are implemented in ‘ae_realization.py’.

To generate a realization, you should normally call the generate_realization method of a type instance.

generate_realization(bottom_surface, top_surface, stratum, grid)[source]

Generates a realization of the AE type.

For example, if the AE geometry is ‘trunc_ellip’, this generates a TruncEllip object, i.e. a sub-layer containing trough features within a stratum.

Parameters

  • bottom_surface (ContactSurface object) – The bottom surface of the AE.

  • top_surface (ContactSurface object) – The top surface of the AE.

  • model (Model object) –

hyvr.geo.channel module

class hyvr.geo.channel.Channel

Bases: object

Channel implementation.

This is work in progress, as we currently lack a centerline construction method.

a
azim
cos_dip
depth
dip
dipsets
dont_check
facies
facies_array
lag_facies
lag_height
layer_dist
len_centerline
min_dx_dy
num_facies
num_ha
shift
sin_dip
vx
vy
width
x_center
y_center
zmax
zmin
ztop

hyvr.geo.channel_ae module

class hyvr.geo.channel_ae.ChannelAE

Bases: hyvr.geo.ae_realization.AERealization

create_object_arrays()

This method is called after the object list has been created, so after all objects have been generated.

Its main purpose is to create fixed size arrays of object attributes as members of the AERealization subtype, so that they can be accessed without needing to access the python list of objects.

All subclasses of AERealization must override this method.

generate_objects()

Generate channel objects and place them in the domain.

The following was just an idea I had, ignore it.

The channels follow a mean flow angle that is given in the parameter file. From this angle, a flow axis is constructed as the “center axis” in direction of the flow angle. (This means the flow axis is chosen such that the projection of the (x-y-)domain onto the axis perpendicular to the flow axis is cut in half by the flow axis.)

The channels are then created as random 1-d functions of the position on the flow axis, with random starting points (within the projection of the domain onto the axis perpendicular to the flow axis). This can then be rotated to get an approximation of the channel curve (a set of points which lie on the curve).

On every level, ‘channel_no’ channels are constructed. They share the same flow angle but are shifted w.r.t. the flow axis.

Its also possible to let the channels migrate in time, i.e. slightly change the flow angle and the distance between the channels (by changing their starting points).

maybe_assign_points_to_object()

This function checks whether the current grid cell with given coordinates is inside the trough and assigns facies, azimuth and dip by altering the passed arrays.

Parameters

  • oi (int) – object index

  • geo_ids (np.int32 array of size 4) – This holds the geological indices, i.e. facies number, architectural element number, hydrofacies assemblage (ha) and hydrofacies assemblage type (hat).

  • angles (double array) – array of size 2 that holds the azimuth and dip of the cell. This will also be altered.

  • y, z (x,) – cell coordinates

  • y_idx (x_idx,) – indices of x and y position in grid.

  • grid (Grid object) –

object_a
object_cos_dip
object_depth
object_dipsets
object_dont_check
object_lag_facies
object_lag_height
object_layer_dist
object_len_centerline
object_min_dx_dy
object_shift
object_sin_dip
object_vx
object_vy
object_width
object_x_center
object_y_center
object_ztop

hyvr.geo.channel_utils module

hyvr.geo.channel_utils.ferguson_curve(grid, h, k, ds, eps_factor, flow_angle, domain_length, xstart, ystart, width)[source]

Simulate extruded parabola centrelines using the Ferguson (1976) disturbed meander model Implementation of AR2 autoregressive model http://onlinelibrary.wiley.com/doi/10.1002/esp.3290010403/full

Parameters

  • grid (Grid object) –

  • h (float) – Height (Ferguson model parameter)

  • k (float) – Wave number (Ferguson model parameter)

  • ds (float) – Curve distance for calculations (Ferguson model parameter)

  • eps_factor (float) – Random background noise (Ferguson model parameter)

  • flow_angle (float) – Angle of mean flow direction, in radians

  • domain_length (float) – Length of the domain in mean flow direction

  • ystart (xstart,) – Starting coordinates of the channel centerline

  • width (float) – Width of the channel

Returns

outputs – Simulated extruded parabola centerlines: storage array containing values for x coordinate, y coordinate, vx and vy

Return type

float matrix

hyvr.geo.channel_utils.ferguson_theta(s, eps_factor, k, h)[source]

Calculate curve direction angle

Parameters

  • s – Steps within generated curve distance

  • eps_factor – Random background noise

  • k – Wave number

  • h – Height

Returns

th_store (array) - Curve direction angle

hyvr.geo.channel_utils.thetaAR2(t1, t2, k, h, eps)[source]

Implementation of AR2 autoregressive model (Ferguson, 1976, Eq.15) http://onlinelibrary.wiley.com/doi/10.1002/esp.3290010403/full

Parameters

  • t1 – theta(i-1)

  • t2 – theta(i-2)

  • k – Wavenumber

  • h – Height

  • eps – Random background noise

Returns

2nd-order autoregression (AR2)

hyvr.geo.contact_surface module

class hyvr.geo.contact_surface.ContactSurface

Bases: object

nx
ny
surface
use_higher_surface_value()
use_lower_surface_value()
z
zmax
zmean
zmin

hyvr.geo.contact_surface_utils module

hyvr.geo.contact_surface_utils.parse_contact_model(param_list, depth=False)[source]

hyvr.geo.grid module

hyvr.geo.model module

This class holds the abstraction of a HyVR model. It can be broadly structured in two parts:

  • Model description: This consists of all info about the model that is read from the ini-file.

  • Model realization: This consists of several 3-d arrays which hold the

    data for the different variables on the grid. This is what get’s put out in the end. The data is stored in a dictionary of arrays.

class hyvr.geo.model.Model(model_dict, strata_dict, elements, flowtrans_section)[source]

Bases: object

generate_ae_types(elements)[source]

Read in elements sections and generate type objects

generate_hydraulic_parameters(hydraulics)[source]

Generates (heterogeneous) hydraulic parameters

Parameters

hydraulics (dictionary) – parsed hydraulics section of the ini-file

generate_model()[source]

Generates the model.

This function generates the full model by creating all strata, the AEs within the strata and the geometrical objects within these.

The method starts from the domain top and calls the strata generation function for each stratum, which handles the generation of architectural elements.

print_model()[source]

Prints a long description of the model. This is mainly for debugging.

hyvr.geo.sheet module

class hyvr.geo.sheet.Sheet

Bases: object

This class holds parameters for single sheet objects.

azim
bottom_surface
dip
dipsets
facies
facies_array
layer_dist
normvec_x
normvec_y
normvec_z
num_facies
num_ha
shift
top_surface
zmax
zmin

hyvr.geo.sheet_ae module

class hyvr.geo.sheet_ae.SheetAE

Bases: hyvr.geo.ae_realization.AERealization

create_object_arrays()

This method is called after the object list has been created, so after all objects have been generated.

Its main purpose is to create fixed size arrays of object attributes as members of the AERealization subtype, so that they can be accessed without needing to access the python list of objects.

All subclasses of AERealization must override this method.

generate_objects()

Generate sheet objects and place them in the domain.

maybe_assign_points_to_object()

This function checks whether the current grid cell with given coordinates is inside the trough and assigns facies, azimuth and dip by altering the passed arrays.

Parameters

  • oi (int) – object index

  • geo_ids (np.int32 array of size 4) – This holds the geological indices, i.e. facies number, architectural element number, hydrofacies assemblage (ha) and hydrofacies assemblage type (hat).

  • angles (double array) – array of size 2 that holds the azimuth and dip of the cell. This will also be altered.

  • y, z (x,) – cell coordinates

  • y_idx (x_idx,) – indices of x and y position in grid.

  • grid (Grid object) –

object_bottom_surface
object_bottom_surface_zmean
object_dipsets
object_layer_dist
object_normvec_x
object_normvec_y
object_normvec_z
object_shift
object_top_surface

hyvr.geo.stratum module

class hyvr.geo.stratum.Stratum(bottom_surface, top_surface, params, ae_types, grid)[source]

Bases: object

hyvr.geo.stratum.prob_choose(types, probs)[source]

hyvr.geo.trough module

class hyvr.geo.trough.Trough

Bases: object

This class holds all parameters of a geometrical trough object.

The only method of this class is __init__, as after creating all trough objects, the list of troughs in TroughAE will be converted to multiple lists containing of attribute values for easier access from Cython.

In principle this could also be implemented in pure Python, anyone interested in rewriting this class (and Channel and Sheet) can do so.

a
alpha
azim
b
c
cosalpha
cosdip
dip
facies
facies_array
lag
lag_facies
layer_dist
max_ab
normvec_x
normvec_y
normvec_z
num_facies
num_ha
shift
sinalpha
structure
x
y
z
zmax
zmin

hyvr.geo.trough_ae module

class hyvr.geo.trough_ae.TroughAE

Bases: hyvr.geo.ae_realization.AERealization

create_object_arrays()

This method is called after the object list has been created, so after all objects have been generated.

Its main purpose is to create fixed size arrays of object attributes as members of the AERealization subtype, so that they can be accessed without needing to access the python list of objects.

All subclasses of AERealization must override this method.

generate_objects()

Generate trough objects and place them in the domain.

maybe_assign_points_to_object()

This function checks whether the current grid cell with given coordinates is inside the trough and assigns facies, azimuth and dip by altering the passed arrays.

Parameters

  • oi (int) – object index

  • geo_ids (np.int32 array of size 4) – This holds the geological indices, i.e. facies number, architectural element number, hydrofacies assemblage (ha) and hydrofacies assemblage type (hat).

  • angles (double array) – array of size 2 that holds the azimuth and dip of the cell. This will also be altered.

  • y, z (x,) – cell coordinates

  • y_idx (x_idx,) – indices of x and y position in grid.

  • grid (Grid object) –

object_a
object_alpha
object_b
object_c
object_cosalpha
object_cosdip
object_lag
object_lag_facies
object_layer_dist
object_max_ab
object_normvec_x
object_normvec_y
object_normvec_z
object_shift
object_sinalpha
object_structure
object_x
object_y
object_z

hyvr.geo.trough_utils module

hyvr.geo.trough_utils.generate_trough_positions(bottom_surface, top_surface, type_params, grid)[source]
hyvr.geo.trough_utils.rand_trough_params(type_params, z, grid)[source]