# Interpolation¶

This treatment interpolates a multiblock source CFD domain on another target CFD domain.

The interpolation method is the Inverse Distance Weighting.

The complexity is N*logN thanks to the use of a fast spatial search structure (kd-tree).

The set of N known data points: $$[ \ldots, (x_i, y_i), \ldots]$$

Interpolated value u at a given point x based on samples $$\displaystyle u_{i}=u(x_{i})$$:

$$\displaystyle u(x) = \frac{\sum_1^N \frac{1}{d(x,x_i)^p} u_i} {\sum_1^N \frac{1}{d(x,x_i)^p}}$$ if $$\displaystyle d(x,x_i) \neq 0 \forall i$$

$$\displaystyle u(x) = u_i$$ if $$\displaystyle \exists i, d(x,x_i) = 0$$

x is the arbitrary interpolated point. $$\displaystyle x_i$$ is the known interpolating point. $$\displaystyle d(x,x_i)$$ is the distance from the known point $$\displaystyle x_i$$ to the unknown point x. N is the total number of known points used in interpolation. p is the power parameter.

Warning

dependency on scipy.spatial

## Input Parameters¶

The following keys can be set:

• source – (type = antares.Base ) – The source mesh (with the variables you want to interpolate).

• target – (type = antares.Base ) – The mesh to interpolate on. The base is modified at the output.

• coordinates – (default = None, type = list(str) ) – The variables that serve as a basis for the interpolation. The KDTree extracts closest points based on these variables.If None, use the coordinates from the source or target base.

• variables – (default = None, type = list(str) ) – List of variables to apply for the interpolation. If None, all variable will be interpolated, except the coordinates. Using this parameter allow to call the interpolation treatment on bases with mixed variable types.

• tolerance – (default = 1e-10, type = float ) – The value that tells if the distance from the closest point is smaller than this tolerance, the interpolated value is equal to the value of that closest point

• duplicated_tolerance – (default = None, type = int ) – Number of decimal places to round to for duplicated points detection. If decimals is negative, it specifies the number of positions to the left of the decimal point. If None the detection will be exact.

• invdist_power – (default = 1, type = int ) – The power of the inverse distance, this acts like a norm (1, 2, or infinity etc.) and can be used to smooth/refined the interpolation.

• nb_points – (default = None, type = int ) – The number of closest points used for the interpolation.If not provided by the user, it will automatically be set depending on the type of cells in the mesh

• with_boundaries – (default = False, type = bool ) – Whether or not to use data from the boundaries.

## Main functions¶

class antares.treatment.TreatmentInterpolation.TreatmentInterpolation

Class for Interpolation Treatment.

The UserInterface derivation allows to have parameter default values

execute()

Interpolate data from a base onto data from another base.

Returns

the target base containing the results

Return type

antares.Base

## Example¶

"""
This example illustrates the interpolation of a
cloud of points against another cloud of points
"""

from copy import copy
import os
if not os.path.isdir('OUTPUT'):
os.makedirs('OUTPUT')

from antares import Reader, Treatment, Writer

# ------------------
# ------------------
reader['filename'] = os.path.join('..', 'data', 'ROTOR37', 'GENERIC', 'flow_<zone>_<instant>.dat')

print(source_base)

# --------------------
# Build a target base
# --------------------
# (same mesh but at cell centers)
tmp_base = copy(source_base[:, :, ('x', 'y', 'z')])
tmp_base.node_to_cell()
target_base = tmp_base.get_location('cell')

# --------------
# Interpolation
# --------------
treatment = Treatment('interpolation')
treatment['source'] = source_base
treatment['target'] = target_base
result = treatment.execute()

# Note that the result contains the same instants as the source and
# each instant contains the variable interpolated and the coordinates
# used for interpolation
print(result)
# >>>  Zone object
#      - instants : ['ite0', 'ite1', 'ite2']
print(result)
# >>>  Instant object
#      - variables : [('x', 'cell'), ('y', 'cell'), ('z', 'cell'), ('ro', 'cell'), ('rovx', 'cell'), ('rovy', 'cell'), ('rovz', 'cell'), ('roE', 'cell')]
#      - shape : None

# -------------------
# Writing the result
# -------------------
writer = Writer('bin_tp')
writer['base'] = result
writer['filename'] = os.path.join('OUTPUT', 'ex_interpolation.plt')
writer.dump()