Source code for eleos.shapes

""" 
Each Shape object represents a differnet model in NEMESIS. Eg. ``Shape1`` represents profile shape 1 which is a constant VMR 
up to a knee pressure then a fractional scale height. For a ful list of possible models see the NEMESIS manual. Please note: 
due to the large number of profile shapes NEMESIS supports, 

Adding a Shape subclass
=======================

All subclasses of Shape must implement:
    - ID: Integer ID of the shape in NEMESIS
    - CONSTANTS: The names of parameters that are not fitted by NEMESIS but required for the model
    - VARIABLES: The names of the parameters in order they appear in the mre file (do not include error parameters here). **These MUST be defined in the order that they appear in the .apr file!!**
    - generate_apr_data: Returns a string of the profile shape parameters in the format
                         that the .apr file expects.

Additionally, some shapes require additional files in the core (eg. tempapr.dat). If this is the case,
override the create_required_files method.

Finally, add the new ShapeN class to the ALL_SHAPES list at the end of the file.

Class template::

    @shapeclass
    class ShapeN(Shape):
        "copy the desciption from the NEMESIS manual here (use triple quotes!)"
        ID: ClassVar[int] = N
        CONSTANTS: ClassVar[list[str]] = ["arg_name_1", ...]
        VARIABLES: ClassVar[list[str]] = ["arg_name_2", "arg_name_3", ...]
        arg_name_1: type_1
        arg_name_2: type_2
        arg_name_2_error: type_2
        ...

        def generate_apr_data(self):
            return a string for apr file

        # optionally
        def create_required_files(self, directory):
            shutil.copy(..., directory)

"""

import numpy as np

from dataclasses import dataclass
from typing import ClassVar
import shutil

from . import parsers
from . import utils


[docs] def shapeclass(*args, **kwargs): """Syntatic sugar for ``@dataclass(repr=False)``""" return dataclass(*args, **kwargs, repr=False)
[docs] class Shape:
[docs] def __init__(self): """Do not instantiate directly, use a subclass""" self.CONSTANTS = [] self.VARIABLES = []
def __repr__(self): return f"<Shape{self.ID}>" def _set_prior_to_retrieved(self, use_retrieval_errors=False): """Set the prior values to the retrieved values and delete the retrived_* attributes.""" for name in self.VARIABLES: setattr(self, name, getattr(self, f"retrieved_{name}")) delattr(self, f"retrieved_{name}") if use_retrieval_errors: setattr(self, f"{name}_error", getattr(self, f"retrieved_{name}_error")) delattr(self, f"retrieved_{name}_error")
[docs] def generate_apr_data(self): """Generate the section of the .apr file that stores the shape parameters Args: None Returns: str: The string to write to the .apr file""" pass
[docs] def create_required_files(self, directory): """Some Shapes require additional files to be created/copied into the core directory""" pass
[docs] def sample_from_distribution(self, **kwargs): """ Set parameters of the Shape to random values pulled from a normal distribution, with arguments mean_[parameter] and std_[parameter]. For example: calling Shape4.sample_from_distribution(mean_deep_vmr=0.1, std_deep_vmr=0.05) will set the deep vmr to a random value drawn from a gaussian with the centre at 0.1 and a standard deviation of 0.05. Multiple parameters may be specified at once. Args: * mean_[parameter]: The mean of the distribution to sample for parameter [parameter] * std_[parameter]: The standard deviation of the distribution to sample for parameter [parameter] Returns: None """ for param in kwargs.keys(): for vp in self.CONSTANTS + self.VARIABLES: if vp in param: if f"mean_{vp}" in kwargs.keys(): mean = kwargs[f"mean_vp"] else: mean = getattr(self, vp) setattr(self, vp, np.random.normal(mean, kwargs[f"std_{vp}"]))
[docs] def check_validity(self): """Check that the parameters for the Shape are valid. This should be overridden by subclasses""" return True
[docs] @shapeclass class Shape0(Shape): """Profile is to be treated as continuous over the pressure range of runname.ref, the next line of the .apr file should then contain a filename, which specifies the a priori profile as a function of height and should have the same number of levels as the .ref file. In Eleos, one can specify either a filepath or give the values and error to be used directly. Parameters ---------- filepath : str, optional Path to the input profile file. values : list[float], optional Profile values to be written if no file is given. errors : list[float], optional Error values corresponding to `values`. """ ID: ClassVar[int] = 0 CONSTANTS: ClassVar[list[str]] = [] VARIABLES: ClassVar[list[str]] = [] filepath: str = None values: list[float] = None errors: list[float] = None
[docs] def create_required_files(self, directory): if self.filepath is not None: shutil.copy(self.filepath, directory) self.values = [] self.errors = [] with open(self.filepath) as file: lines = file.read().split("\n")[1:] for line in lines: p, v, e = utils.get_floats_from_string(line) self.values.append(v) self.errors.append(e) else: self.filepath = directory / "shape0.dat" ref = parsers.NemesisRef(directory / "nemesis.ref") with open(self.filepath, mode="w+") as file: file.write(f"{len(ref.data.pressure)} 1\n#") for p, v, e in zip(ref.data.pressure, self.values, self.errors): file.write(f"{p:08e} {v:08e} {e:08e}\n")
[docs] def generate_apr_data(self): return self.filepath.name
[docs] @shapeclass class Shape1(Shape): """Profile is to be represented as a deep VMR up to a certain ‘knee’ pressure, and then a defined fractional scale height. Parameters ---------- knee_pressure : float Pressure at which the transition occurs. deep_vmr : float Deep volume mixing ratio. deep_vmr_error : float Error estimate for deep_vmr. fsh : float Fractional scale height. fsh_error : float Error estimate for fsh. """ ID: ClassVar[int] = 1 CONSTANTS: ClassVar[list[str]] = ["knee_pressure"] VARIABLES: ClassVar[list[str]] = ["deep_vmr", "fsh"] knee_pressure: float deep_vmr: float deep_vmr_error: float fsh: float fsh_error: float
[docs] def generate_apr_data(self): return f"{self.knee_pressure}\n{self.deep_vmr} {self.deep_vmr_error}\n{self.fsh} {self.fsh_error}"
[docs] def check_validity(self): return self.deep_vmr >= 0 and self.fsh > 0 and self.knee_pressure > 0
[docs] @shapeclass class Shape2(Shape): """Profile is to be represented by a simple scaling of the corresponding profile runname.ref, aerosol.ref, parah2.ref or fcloud.ref. Parameters ---------- scale_factor : float Scaling factor applied to the reference profile. scale_factor_error : float Error estimate for scale_factor. """ ID: ClassVar[int] = 2 CONSTANTS: ClassVar[list[str]] = [] VARIABLES: ClassVar[list[str]] = ["scale_factor"] scale_factor: float scale_factor_error: float
[docs] def generate_apr_data(self): return f"{self.scale_factor} {self.scale_factor_error}"
[docs] @shapeclass class Shape4(Shape): """Very similar to Shape1, but knee pressure is also a variable parameter. Parameters ---------- knee_pressure : float Pressure at which the transition occurs. knee_pressure_error : float Error estimate for knee_pressure. deep_vmr : float Deep volume mixing ratio. deep_vmr_error : float Error estimate for deep_vmr. fsh : float Fractional scale height. fsh_error : float Error estimate for fsh. """ ID: ClassVar[int] = 4 CONSTANTS: ClassVar[list[str]] = [] VARIABLES: ClassVar[list[str]] = ["deep_vmr", "fsh", "knee_pressure"] knee_pressure: float knee_pressure_error: float deep_vmr: float deep_vmr_error: float fsh: float fsh_error: float
[docs] def generate_apr_data(self): return f"{self.knee_pressure} {self.knee_pressure_error}\n{self.deep_vmr} {self.deep_vmr_error}\n{self.fsh} {self.fsh_error}"
[docs] def check_validity(self): return self.deep_vmr >= 0 and self.fsh > 0 and self.knee_pressure > 0
[docs] @shapeclass class Shape20(Shape): """Similar to Shape1, but profile forced to a small number above tropopause. Parameters ---------- knee_pressure : float Pressure at which the transition occurs. tropopause_pressure : float Pressure representing the tropopause. deep_vmr : float Deep volume mixing ratio. deep_vmr_error : float Error estimate for deep_vmr. fsh : float Fractional scale height. fsh_error : float Error estimate for fsh. """ ID: ClassVar[int] = 20 CONSTANTS: ClassVar[list[str]] = ["knee_pressure", "tropopause_pressure"] VARIABLES: ClassVar[list[str]] = ["deep_vmr", "fsh"] knee_pressure: float tropopause_pressure: float deep_vmr: float deep_vmr_error: float fsh: float fsh_error: float
[docs] def generate_apr_data(self): return f"{self.knee_pressure} {self.tropopause_pressure}\n{self.deep_vmr} {self.deep_vmr_error}\n{self.fsh} {self.fsh_error}"
[docs] def check_validity(self): return self.deep_vmr >= 0 and self.fsh > 0 and self.knee_pressure > 0 and self.tropopause_pressure > 0
[docs] @shapeclass class Shape32(Shape): """Cloud profile with variable base pressure, opacity, and fractional scale height. Parameters ---------- base_pressure : float Cloud base pressure (bar). base_pressure_error : float Error estimate for base_pressure. opacity : float Cloud opacity (log-scaled). opacity_error : float Error estimate for opacity. fsh : float Fractional scale height (log-scaled). fsh_error : float Error estimate for fsh. """ ID: ClassVar[int] = 32 CONSTANTS: ClassVar[list[str]] = [] VARIABLES: ClassVar[list[str]] = ["opacity", "fsh", "base_pressure"] base_pressure: float base_pressure_error: float opacity: float opacity_error: float fsh: float fsh_error: float
[docs] def generate_apr_data(self): return f"{self.base_pressure} {self.base_pressure_error}\n{self.opacity} {self.opacity_error}\n{self.fsh} {self.fsh_error}"
[docs] def check_validity(self): return self.opacity >= 0 and self.fsh > 0 and self.base_pressure > 0
[docs] @shapeclass class Shape37(Shape): """Cloud with constant opacity/bar between two pressure levels. Parameters ---------- bottom_pressure : float Bottom pressure level (bar). top_pressure : float Top pressure level (bar). opacity : float Cloud opacity per bar. opacity_error : float Error estimate for opacity. """ ID: ClassVar[int] = 37 CONSTANTS: ClassVar[list[str]] = ["bottom_pressure", "top_pressure"] VARIABLES: ClassVar[list[str]] = ["opacity"] bottom_pressure: float top_pressure: float opacity: float opacity_error: float
[docs] def generate_apr_data(self): return f"{self.bottom_pressure} {self.top_pressure}\n{self.opacity} {self.opacity_error}"
[docs] def check_validity(self): return self.opacity >= 0 and self.bottom_pressure > 0 and self.top_pressure < self.bottom_pressure
[docs] @shapeclass class Shape47(Shape): """Cloud centred at specified pressure with variable FWHM and opacity. Parameters ---------- central_pressure : float Pressure where cloud distribution peaks (bar). central_pressure_error : float Error estimate for central_pressure (bar). pressure_width : float Full width at half maximum in log pressure units. pressure_width_error : float Error estimate for pressure_width. opacity : float Total cloud opacity. opacity_error : float Error estimate for opacity. """ ID: ClassVar[int] = 47 CONSTANTS: ClassVar[list[str]] = [] VARIABLES: ClassVar[list[str]] = ["opacity", "central_pressure", "pressure_width"] central_pressure: float central_pressure_error: float pressure_width: float pressure_width_error: float opacity: float opacity_error: float
[docs] def generate_apr_data(self): return f"{self.opacity} {self.opacity_error}\n{self.central_pressure} {self.central_pressure_error}\n{self.pressure_width} {self.pressure_width_error}"
[docs] def check_validity(self): return self.opacity >= 0 and self.central_pressure > 0 and self.pressure_width > 0
[docs] @shapeclass class Shape48(Shape): """Cloud profile with variable base, top pressure, opacity, and scale height. Parameters ---------- base_pressure : float Cloud base pressure (bar). base_pressure_error : float Error estimate for base_pressure (bar). top_pressure : float Cloud top pressure (bar). top_pressure_error : float Error estimate for top_pressure (bar). opacity : float Cloud opacity. opacity_error : float Error estimate for opacity. fsh : float Fractional scale height. fsh_error : float Error estimate for fsh. """ ID: ClassVar[int] = 48 CONSTANTS: ClassVar[list[str]] = [] VARIABLES: ClassVar[list[str]] = ["opacity", "fsh", "base_pressure", "top_pressure"] base_pressure: float base_pressure_error: float top_pressure: float top_pressure_error: float opacity: float opacity_error: float fsh: float fsh_error: float
[docs] def generate_apr_data(self): return f"{self.base_pressure} {self.base_pressure_error}\n{self.top_pressure} {self.top_pressure_error}\n{self.opacity} {self.opacity_error}\n{self.fsh} {self.fsh_error}"
[docs] def check_validity(self): return self.opacity >= 0 and self.fsh > 0 and self.base_pressure > 0 and self.top_pressure < self.base_pressure
[docs] def get_shape_from_id(id_): """Given a shape ID integer, return a reference to the class corresponding to that ID. Note that this returns a class, not an instantiated object. Args: id_: Shape ID to get Returns: Shape: Class object of the corresponding shape""" id_ = int(id_) for shape_class in ALL_SHAPES: if shape_class.ID == id_: return shape_class
ALL_SHAPES = [Shape0, Shape1, Shape2, Shape4, Shape20, Shape32, Shape37, Shape47, Shape48] #: A list of all the Shape classes