IR-e0dZddlZddlmZddlmZ ddlm Z dZ n #e $rdZ YnwxYwgd Z d d giZ Gd d eZddZeddZeddZdZdeze_deze_dS)a Tabular models. Tabular models of any dimension can be created using `tabular_model`. For convenience `Tabular1D` and `Tabular2D` are provided. Examples -------- >>> table = np.array([[ 3., 0., 0.], ... [ 0., 2., 0.], ... [ 0., 0., 0.]]) >>> points = ([1, 2, 3], [1, 2, 3]) >>> t2 = Tabular2D(points, lookup_table=table, bounds_error=False, ... fill_value=None, method='nearest') N)units)Model)interpnTF) tabular_model Tabular1D Tabular2DrscipyceZdZdZdZdZdZdZdZdddde j ffd Z dZ d Z ed Zed Zed Zd ZedZxZS)_Tabulara Returns an interpolated lookup table value. Parameters ---------- points : tuple of ndarray of float, optional The points defining the regular grid in n dimensions. ndarray must have shapes (m1, ), ..., (mn, ), lookup_table : array-like The data on a regular grid in n dimensions. Must have shapes (m1, ..., mn, ...) method : str, optional The method of interpolation to perform. Supported are "linear" and "nearest", and "splinef2d". "splinef2d" is only supported for 2-dimensional data. Default is "linear". bounds_error : bool, optional If True, when interpolated values are requested outside of the domain of the input data, a ValueError is raised. If False, then ``fill_value`` is used. fill_value : float or `~astropy.units.Quantity`, optional If provided, the value to use for points outside of the interpolation domain. If None, values outside the domain are extrapolated. Extrapolation is not supported by method "splinef2d". If Quantity is given, it will be converted to the unit of ``lookup_table``, if applicable. Returns ------- value : ndarray Interpolated values at input coordinates. Raises ------ ImportError Scipy is not installed. Notes ----- Uses `scipy.interpolate.interpn`. FTrNlinearc 2|dd}|dkrtdtjdi|d|_|t dt |tjstj |}|j j |j krt d|j j d|td|jD}n|j dkrt |ts|f}t|}||j krt d |j d |d |dkrLt |d tjr,td |Ddkrt dt |tjrQt |tjst d|jd||jj}||_||_ ||_||_||_dS)Nn_modelsrzOnly n_models=1 is supported.)yzMust provide a lookup table.z%lookup_table should be an array with z dimensions.c3LK|]}tj|tV dS))dtypeN)nparangefloat).0xs 8lib/python3.11/site-packages/astropy/modeling/tabular.py z$_Tabular.__init__..vs1QQ29Qe444QQQQQQzExpected grid points in z directions, got .rc0h|]}t|ddS)unitN)getattrrps r z$_Tabular.__init__..s$BBBaFD11BBBrz#points must all have the same unit.zfill value is in z but expected to be unitless.)getNotImplementedErrorsuper__init__outputs ValueError isinstanceuQuantityrasarray lookup_tablendimtupleshapelenrtovaluepoints bounds_errormethod fill_value) selfr4r-r6r5r7kwargsrnpts __class__s rr&z_Tabular.__init__[sT::j!,, a<<%&EFF F""6"""  ;<< <, 33 4:l33L   !\%6 6 68$)888  >QQl>PQQQQQFF A%%j.G.G% v;;D|((( C#(CC;?CCC qvay!*55BB6BBBCCaGG !FGGG j!* - - @lAJ77  V VVV$|'899?J (( $rcBd|jjd|jd|jdS)N)r;__name__r4r-r8s r__repr__z_Tabular.__repr__s@ 2' 2 2 2 2 - 2 2 2 rc d|jjfd|jfd|jfd|jfdd|jfd|jfd|jfd |jfd |j fg }d |D}d |S) NrNameN_inputs N_outputs) Parametersz pointsz lookup_tablez methodz fill_valuez bounds_errorc&g|]\}}||d|S)Nz: r")rkeywordr3s r z$_Tabular.__str__..s:     ! !% ! !   r ) r;r>namen_inputs n_outputsr4r-r6r7r5join)r8default_keywordspartss r__str__z_Tabular.__str__s dn- . TY   ' $. )   % t0 1  % T_ - t0 1    "2    yyrc||jdttjsdSfd|jDS)Nrc i|] }|j Sr")r)rrptss r z(_Tabular.input_units..s11138111r)r4r)r*r+inputs)r8rTs @r input_unitsz_Tabular.input_unitssCk!n#qz** 41111T[1111rcvt|jtjsdS|jd|jjiS)Nr)r)r-r*r+r'rr?s r return_unitsz_Tabular.return_unitss5$+QZ88 4 Q!2!788rcpd|jDddd}t|dkr|d}|S)ay Tuple defining the default ``bounding_box`` limits, ``(points_low, points_high)``. Examples -------- >>> from astropy.modeling.models import Tabular1D, Tabular2D >>> t1 = Tabular1D(points=[1, 2, 3], lookup_table=[10, 20, 30]) >>> t1.bounding_box ModelBoundingBox( intervals={ x: Interval(lower=1, upper=3) } model=Tabular1D(inputs=('x',)) order='C' ) >>> t2 = Tabular2D(points=[[1, 2, 3], [2, 3, 4]], ... lookup_table=[[10, 20, 30], [20, 30, 40]]) >>> t2.bounding_box ModelBoundingBox( intervals={ x: Interval(lower=1, upper=3) y: Interval(lower=2, upper=4) } model=Tabular2D(inputs=('x', 'y')) order='C' ) cJg|] }t|t|f!Sr")minmaxrs rrIz)_Tabular.bounding_box..s)666QQQ 666rNrr)r4r1)r8bboxs r bounding_boxz_Tabular.bounding_boxsA>76$+666ttt< t99>>7D rc.tj|}|djd|d|jD}tj|j}t stdt|j |j ||j |j |j }t|j tjr4t|j dtjs||j jz}|jdkr|}nfd|D}|S)a Return the interpolated values at the input coordinates. Parameters ---------- inputs : list of scalar or list of ndarray Input coordinates. The number of inputs must be equal to the dimensions of the lookup table. rc6g|]}|Sr")flatten)rinps rrIz%_Tabular.evaluate..s CCCC#++--CCCrNzTabular model requires scipy.)r6r5r7rc:g|]}|Sr")reshape)rrr0s rrIz%_Tabular.evaluate..s%7771aii&&777r)rbroadcast_arraysr0rLarrayT has_scipy ImportErrorrr4r-r6r5r7r)r*r+rrMrf)r8rVresultr0s @revaluatez_Tabular.evaluates$f-q CC6/DM/+BCCC&!!# ?=>> > K   ;*     d' 4 4 5Z KNAJ> >  5d/44F >Q  ^^E**FF7777777F rc|jdkrtjtj|jdkr|j}|jd}n]tjtj|jdkr'|jddd}|jdddd}nt t|||j|j |j St d)Nrrr^)r4r-r6r5r7zHAn analytical inverse transform has not been implemented for this model.) rLralldiffr-r4r$rr6r5r7)r8r4r-s rinversez_Tabular.inverses =A  vbgd/001455 **#{1~  122Q677 **44R40#{1~ddd3 *)){!.?  " V   r)r> __module__ __qualname____doc__r fittablestandard_broadcasting _is_dynamic_idrnanr&r@rQpropertyrWrYr`rnrr __classcell__)r;s@rr r 's((TFH!K C6 7%7%7%7%7%7%r      ,22X2 99X9 !!X!F$$$L  X     rr c4|dkrtdtjdg|z}||dd}|dkrd|d<nd|d<|&tj}txjdz c_d |}t t |tf|}d |_|S) a Make a ``Tabular`` model where ``n_inputs`` is based on the dimension of the lookup_table. This model has to be further initialized and when evaluated returns the interpolated values. Parameters ---------- dim : int Dimensions of the lookup table. name : str Name for the class. Examples -------- >>> table = np.array([[3., 0., 0.], ... [0., 2., 0.], ... [0., 0., 0.]]) >>> tab = tabular_model(2, name='Tabular2D') >>> print(tab) Name: Tabular2D N_inputs: 2 N_outputs: 1 >>> points = ([1, 2, 3], [1, 2, 3]) Setting fill_value to None, allows extrapolation. >>> m = tab(points, lookup_table=table, name='my_table', ... bounds_error=False, fill_value=None, method='nearest') >>> xinterp = [0, 1, 1.5, 2.72, 3.14] >>> m(xinterp, xinterp) # doctest: +FLOAT_CMP array([3., 3., 3., 0., 0.]) rz.Lookup table must have at least one dimension.)r-rLrMT _separableFNTabularzastropy.modeling.tabular)r(rzerosr rytypestrrs)dimrKtablemembersmodel_id model_classs rrr$sN QwwIJJJ HaS3Y  E$#AFFG axx $  %  |<  ###s4yy8+w77K7K rr)rKr~r a method : str, optional The method of interpolation to perform. Supported are "linear" and "nearest", and "splinef2d". "splinef2d" is only supported for 2-dimensional data. Default is "linear". bounds_error : bool, optional If True, when interpolated values are requested outside of the domain of the input data, a ValueError is raised. If False, then ``fill_value`` is used. fill_value : float, optional If provided, the value to use for points outside of the interpolation domain. If None, values outside the domain are extrapolated. Extrapolation is not supported by method "splinef2d". Returns ------- value : ndarray Interpolated values at input coordinates. Raises ------ ImportError Scipy is not installed. Notes ----- Uses `scipy.interpolate.interpn`. a! Tabular model in 1D. Returns an interpolated lookup table value. Parameters ---------- points : array-like of float of ndim=1. The points defining the regular grid in n dimensions. lookup_table : array-like, of ndim=1. The data in one dimensions. ai Tabular model in 2D. Returns an interpolated lookup table value. Parameters ---------- points : tuple of ndarray of float, optional The points defining the regular grid in n dimensions. ndarray with shapes (m1, m2). lookup_table : array-like The data on a regular grid in 2 dimensions. Shape (m1, m2). )N)runumpyrastropyrr*corerscipy.interpolaterrkrl__all____doctest_requires__r rrr _tab_docsr"rrrsJ$))))))IIIII 6 5 5''3z z z z z uz z z z9999x M!+ . . . M!+ . . .   >     s ''