+ Zjf^RIHt^RIHtHt^RIHt^RIHtH t H t H t H t H t HtHtHtHtHtHtHtHtHtHt^RIt^RIHtHt^RIHtR.t!R R4tR t R#) ) GenericAlias)linalgspecial)check_random_state)asarray atleast_2dreshapezerosnewaxisexppisqrtravelpower atleast_1dsqueezesum transposeonescovN)gaussian_kernel_estimategaussian_kernel_estimate_log)multivariate_normal gaussian_kdeca]tRt^&toRt]!]4tRRltRt ] t Rt Rt RRR/Rllt R tRR ltR tR t]tR ]nRRltRt]R4tRtRtRt]R4t]R4tV3RltRtVtR#)rayRepresentation of a kernel-density estimate using Gaussian kernels. Kernel density estimation is a way to estimate the probability density function (PDF) of a random variable in a non-parametric way. `gaussian_kde` works for both uni-variate and multi-variate data. It includes automatic bandwidth determination. The estimation works best for a unimodal distribution; bimodal or multi-modal distributions tend to be oversmoothed. Parameters ---------- dataset : array_like Datapoints to estimate from. In case of univariate data this is a 1-D array, otherwise a 2-D array with shape (# of dims, # of data). bw_method : str, scalar or callable, optional The method used to calculate the bandwidth factor. This can be 'scott', 'silverman', a scalar constant or a callable. If a scalar, this will be used directly as `factor`. If a callable, it should take a `gaussian_kde` instance as only parameter and return a scalar. If None (default), 'scott' is used. See Notes for more details. weights : array_like, optional weights of datapoints. This must be the same shape as dataset. If None (default), the samples are assumed to be equally weighted Attributes ---------- dataset : ndarray The dataset with which `gaussian_kde` was initialized. d : int Number of dimensions. n : int Number of datapoints. neff : int Effective number of datapoints. .. versionadded:: 1.2.0 factor : float The bandwidth factor obtained from `covariance_factor`. covariance : ndarray The kernel covariance matrix; this is the data covariance matrix multiplied by the square of the bandwidth factor, e.g. ``np.cov(dataset) * factor**2``. inv_cov : ndarray The inverse of `covariance`. Methods ------- evaluate __call__ integrate_gaussian integrate_box_1d integrate_box integrate_kde pdf logpdf resample set_bandwidth covariance_factor marginal Notes ----- Bandwidth selection strongly influences the estimate obtained from the KDE (much more so than the actual shape of the kernel). Bandwidth selection can be done by a "rule of thumb", by cross-validation, by "plug-in methods" or by other means; see [3]_, [4]_ for reviews. `gaussian_kde` uses a rule of thumb, the default is Scott's Rule. Scott's Rule [1]_, implemented as `scotts_factor`, is:: n**(-1./(d+4)), with ``n`` the number of data points and ``d`` the number of dimensions. In the case of unequally weighted points, `scotts_factor` becomes:: neff**(-1./(d+4)), with ``neff`` the effective number of datapoints. Silverman's suggestion for *multivariate* data [2]_, implemented as `silverman_factor`, is:: (n * (d + 2) / 4.)**(-1. / (d + 4)). or in the case of unequally weighted points:: (neff * (d + 2) / 4.)**(-1. / (d + 4)). Note that this is not the same as "Silverman's rule of thumb" [6]_, which may be more robust in the univariate case; see documentation of the ``set_bandwidth`` method for implementing a custom bandwidth rule. Good general descriptions of kernel density estimation can be found in [1]_ and [2]_, the mathematics for this multi-dimensional implementation can be found in [1]_. With a set of weighted samples, the effective number of datapoints ``neff`` is defined by:: neff = sum(weights)^2 / sum(weights^2) as detailed in [5]_. `gaussian_kde` does not currently support data that lies in a lower-dimensional subspace of the space in which it is expressed. For such data, consider performing principal component analysis / dimensionality reduction and using `gaussian_kde` with the transformed data. References ---------- .. [1] D.W. Scott, "Multivariate Density Estimation: Theory, Practice, and Visualization", John Wiley & Sons, New York, Chicester, 1992. .. [2] B.W. Silverman, "Density Estimation for Statistics and Data Analysis", Vol. 26, Monographs on Statistics and Applied Probability, Chapman and Hall, London, 1986. .. [3] B.A. Turlach, "Bandwidth Selection in Kernel Density Estimation: A Review", CORE and Institut de Statistique, Vol. 19, pp. 1-33, 1993. .. [4] D.M. Bashtannyk and R.J. Hyndman, "Bandwidth selection for kernel conditional density estimation", Computational Statistics & Data Analysis, Vol. 36, pp. 279-298, 2001. .. [5] Gray P. G., 1969, Journal of the Royal Statistical Society. Series A (General), 132, 272 .. [6] Kernel density estimation. *Wikipedia.* https://en.wikipedia.org/wiki/Kernel_density_estimation Examples -------- Generate some random two-dimensional data: >>> import numpy as np >>> from scipy import stats >>> def measure(n): ... "Measurement model, return two coupled measurements." ... m1 = np.random.normal(size=n) ... m2 = np.random.normal(scale=0.5, size=n) ... return m1+m2, m1-m2 >>> m1, m2 = measure(2000) >>> xmin = m1.min() >>> xmax = m1.max() >>> ymin = m2.min() >>> ymax = m2.max() Perform a kernel density estimate on the data: >>> X, Y = np.mgrid[xmin:xmax:100j, ymin:ymax:100j] >>> positions = np.vstack([X.ravel(), Y.ravel()]) >>> values = np.vstack([m1, m2]) >>> kernel = stats.gaussian_kde(values) >>> Z = np.reshape(kernel(positions).T, X.shape) Plot the results: >>> import matplotlib.pyplot as plt >>> fig, ax = plt.subplots() >>> ax.imshow(np.rot90(Z), cmap=plt.cm.gist_earth_r, ... extent=[xmin, xmax, ymin, ymax]) >>> ax.plot(m1, m2, 'k.', markersize=2) >>> ax.set_xlim([xmin, xmax]) >>> ax.set_ylim([ymin, ymax]) >>> plt.show() Compare against manual KDE at a point: >>> point = [1, 2] >>> mean = values.T >>> cov = kernel.factor**2 * np.cov(values) >>> X = stats.multivariate_normal(cov=cov) >>> res = kernel.pdf(point) >>> ref = X.pdf(point - mean).sum() / len(mean) >>> np.allclose(res, ref) True Ncj\\V44VnVPP^8g \ R4hVPP wVnVnVe\V4P\4Vn V;P\VP4,un VPP^8wd \ R4h\VP4VP8wd \ R4h^\ P"!VPVP4, VnVP VP8dRp\ V4hVP'VR7R# \(P*dpRp\(P*!T4ThRp?ii;i)z.`dataset` input should have multiple elements.Nz*`weights` input should be one-dimensional.z%`weights` input should be of length na1Number of dimensions is greater than number of samples. This results in a singular data covariance matrix, which cannot be treated using the algorithms implemented in `gaussian_kde`. Note that `gaussian_kde` interprets each *column* of `dataset` to be a point; consider transposing the input to `dataset`. bw_methodabThe data appears to lie in a lower-dimensional subspace of the space in which it is expressed. This has resulted in a singular data covariance matrix, which cannot be treated using the algorithms implemented in `gaussian_kde`. Consider performing principal component analysis / dimensionality reduction and using `gaussian_kde` with the transformed data.)rrdatasetsize ValueErrorshapednrastypefloat_weightsrweightsndimlennpvecdot_neff set_bandwidthr LinAlgError)selfr rr)msges&&&& Q/home/luca/screen-checking/.venv/lib/python3.14/site-packages/scipy/stats/_kde.py__init__gaussian_kde.__init__sF!''"23 ||  1$MN N++  &w/66u=DM MMS/ /M||  A% !MNN4==!TVV+ !HII299T]]DMMBBDJ 66DFF?-C S/ ! 1     3!! 1?C$$S)q 0 1s+E??F2F--F2c \\V44pVPwr#W P8wdOV^8Xd+W0P8Xd\ WP^34p^pMRV RVP 2p\ V4h\ VPV4wrV\V,!VPPVPR,VPVPV4pVR,#)aEvaluate the estimated pdf on a set of points. Parameters ---------- points : (# of dimensions, # of points)-array Alternatively, a (# of dimensions,) vector can be passed in and treated as a single point. Returns ------- values : (# of points,)-array The values at each point. Raises ------ ValueError : if the dimensionality of the input points is different than the dimensionality of the KDE. points have dimension , dataset has dimension NNNNr;r) rrr#r$r r"_get_output_dtype covariancerr Tr)cho_cov)r1pointsr$mr2 output_dtypespecresults&& r4evaluategaussian_kde.evaluates(GFO,|| ;Av!vv+ &&!5/s3004x9 o%.tG )$/ LLNNDLL1 HHdllL2d|c `\\V44p\V4pVPVP38wd\ RVP 24hVPVPVP38wd\ RVP 24hVR\ 3,pVPV,p\P!V4pVPV, p\P!WE4p\P!\P!V^,44p\^\ ,VP^,R, 4V,p\P"!WV^R7R, p \P"!\%V )4VP&^R7V, p V #)a Multiply estimated density by a multivariate Gaussian and integrate over the whole space. Parameters ---------- mean : aray_like A 1-D array, specifying the mean of the Gaussian. cov : array_like A 2-D array, specifying the covariance matrix of the Gaussian. Returns ------- result : scalar The value of the integral. Raises ------ ValueError If the mean or covariance of the input Gaussian differs from the KDE's dimensionality. zmean does not have dimension z#covariance does not have dimension r;@axis)rrrr#r$r"r r>r cho_factorr cho_solver,proddiagonalrr r-r r)) r1meanrsum_cov sum_cov_choldifftdiffsqrt_det norm_constenergiesrEs &&& r4integrate_gaussiangaussian_kde.integrate_gaussian's;0'$-(o ::$&& "4<r rndtrr,r-r))r1lowhighstdevnormalized_lownormalized_highdeltavalues&&& r4integrate_box_1dgaussian_kde.integrate_box_1d\s, 66Q;FG Gd4??+,Q/ll 2e;< !4 => _- ^0LL $,,. rHrngc WPP, W PP, re\P!WeVPVVR7p\ P !WpPRR7#)aComputes the integral of a pdf over a rectangular interval. Parameters ---------- low_bounds : array_like A 1-D array containing the lower bounds of integration. high_bounds : array_like A 1-D array containing the upper bounds of integration. maxpts : int, optional The maximum number of points to use for integration. rng : `numpy.random.Generator`, optional Pseudorandom number generator state. When `rng` is None, a new generator is created using entropy from the operating system. Types other than `numpy.random.Generator` are passed to `numpy.random.default_rng` to instantiate a ``Generator``. Returns ------- value : scalar The result of the integral. ) lower_limitrmaxptsrfrK)r r?rcdfr>r,r-r))r1 low_bounds high_boundsrirfr]r^valuess&&&&$ r4 integrate_boxgaussian_kde.integrate_box~sV./||~~1MT$(( tv yyB77rHc vVPVP8wd \R4hVPVP8dTpTpMTpTpVPVP,p\P !V4pRp\ VP4FpVPRV\3,pVPV, p \P!WY4p \P!W^R7R, p V\P!\V )4VP^R7VPV,,, pK \P!\P!V^,44p \!^\",VP$^,R, 4V ,p Wm,pV#)a' Computes the integral of the product of this kernel density estimate with another. Parameters ---------- other : gaussian_kde instance The other kde. Returns ------- value : scalar The result of the integral. Raises ------ ValueError If the KDEs have different dimensionality. z$KDEs are not the same dimensionalitygr;rKrJ)r$r"r%r>rrMranger r rNr,r-r r)rOrPrr r#)r1othersmalllargerRrSrEirQrTrUrXrVrWs&& r4 integrate_kdegaussian_kde.integrate_kdesG* 77dff CD D 77TVV EEEE""U%5%55((1 uwwA==Aw/D==4'D$$\8Eyy15;H biiXI AFu}}UVGWW WF  772;;|A781r67==#3c#9:XE  rHc RVf\VP4p\V4p\VP \ VP 3\4VPVR74pVPVPWPR7pVPRV3,pWd,#)aRandomly sample a dataset from the estimated pdf. Parameters ---------- size : int, optional The number of samples to draw. If not provided, then the size is the same as the effective number of samples in the underlying dataset. seed : {None, int, `numpy.random.Generator`, `numpy.random.RandomState`}, optional If `seed` is None (or `np.random`), the `numpy.random.RandomState` singleton is used. If `seed` is an int, a new ``RandomState`` instance is used, seeded with `seed`. If `seed` is already a ``Generator`` or ``RandomState`` instance then that instance is used. Returns ------- resample : (self.d, `size`) ndarray The sampled dataset. )r!)r!pr;) intneffrrrr r$r'r>choicer%r)r )r1r!seed random_statenormindicesmeanss&&& r4resamplegaussian_kde.resamples. <tyy>D)$/ 99 466)U #T__4:  %%dff4<<%H QZ(|rHc ^\VPRVP^,, 4#)zGCompute Scott's factor. Returns ------- s : float Scott's factor. rr|r$r1s&r4 scotts_factorgaussian_kde.scotts_factors!TYYTVVAX//rHc \VPVPR,,R, RVP^,, 4#)zSCompute the Silverman factor. Returns ------- s : float The silverman factor. rJg@rrrs&r4silverman_factorgaussian_kde.silverman_factors3TYYs +C/dffQh@@rHa4Computes the bandwidth factor `factor`. The default is `scotts_factor`. A subclass can overwrite this method to provide a different method, or set it through a call to `set_bandwidth`. Returns ------- factor : float The bandwidth factor. c aaSfMSR8XdSPSnMSR8XdSPSnMv\P!S4'd*\ S\ 4'gRSnV3RlSnM1\S4'dSSnV3RlSnM Rp\V4hSP4R#)a*Compute the bandwidth factor with given method. The new bandwidth calculated after a call to `set_bandwidth` is used for subsequent evaluations of the estimated density. Parameters ---------- bw_method : str, scalar or callable, optional The method used to calculate the bandwidth factor. This can be 'scott', 'silverman', a scalar constant or a callable. If a scalar, this will be used directly as `factor`. If a callable, it should take a `gaussian_kde` instance as only parameter and return a scalar. If None (default), nothing happens; the current `covariance_factor` method is kept. Notes ----- .. versionadded:: 0.11 Examples -------- >>> import numpy as np >>> import scipy.stats as stats >>> x1 = np.array([-7, -5, 1, 4, 5.]) >>> kde = stats.gaussian_kde(x1) >>> xs = np.linspace(-10, 10, num=50) >>> y1 = kde(xs) >>> kde.set_bandwidth(bw_method='silverman') >>> y2 = kde(xs) >>> kde.set_bandwidth(bw_method=kde.factor / 3.) >>> y3 = kde(xs) >>> import matplotlib.pyplot as plt >>> fig, ax = plt.subplots() >>> ax.plot(x1, np.full(x1.shape, 1 / (4. * x1.size)), 'bo', ... label='Data points (rescaled)') >>> ax.plot(xs, y1, label='Scott (default)') >>> ax.plot(xs, y2, label='Silverman') >>> ax.plot(xs, y3, label='Const (1/3 * Silverman)') >>> ax.legend() >>> plt.show() Nscott silvermanz use constantc<S#Nrsr4,gaussian_kde.set_bandwidth..FsYrHc&<SPS4#r) _bw_methodrsr4rrIsT__T-BrHzC`bw_method` should be 'scott', 'silverman', a scalar or a callable.) rcovariance_factorrr,isscalar isinstancestrrcallabler"_compute_covariance)r1rr2sff r4r/gaussian_kde.set_bandwidthsX    ' !%)%7%7D " + %%)%:%:D " [[ # #Jy#,F,F,DO%6D " i 'DO%BD "#CS/ !   "rHc  VP4Vn\VR4'gY\\ VP ^RVP R74Vn\P!VPRR7Vn VPVP^,,Vn VPVP,P\P4Vn^\P !\P"!VP\P$!^\&,4,44P)4,VnR#)zSComputes the covariance matrix for each Gaussian kernel using covariance_factor(). _data_cho_covFrowvarbiasaweightsT)lowerN)rfactorhasattrrrr r)_data_covariancercholeskyrr>r&r,float64r@logdiagrr rlog_detrs&r4r gaussian_kde._compute_covarianceQs,,. t_--$.s4<<498< 0F%GD !"(1F1F7;"=D //$++q.@**T[[8@@L  *,''!B$-)8!9::=#%@ rHc VP4Vn\\VP^RVP R74Vn\P!VP 4VP^,, #)rFr) rrrrr r)rrinvrs&r4inv_covgaussian_kde.inv_covcs],,. *3t||A05 ,N!Ozz$//04;;>AArHc $VPV4#)a3 Evaluate the estimated pdf on a provided set of points. Parameters ---------- x : array_like Points at which to evaluate the pdf. Returns ------- pdf : ndarray The pdf evaluated at `x`. Notes ----- This is an alias for `gaussian_kde.evaluate`. See the ``evaluate`` docstring for more details. )rF)r1xs&&r4pdfgaussian_kde.pdfos(}}QrHc \V4pVPwr4W0P8wdOV^8Xd+W@P8Xd\W P^34p^pMRV RVP 2p\ V4h\ VP V4wrg\V,!VPPVPR,VPVPV4pVR,#)z Evaluate the log of the estimated pdf on a provided set of points. Parameters ---------- x : array_like Points at which to evaluate the log-pdf. Returns ------- logpdf : ndarray The log-pdf evaluated at `x`. r8r9r:r<) rr#r$r r"r=r>rr r?r)r@) r1rrAr$rBr2rCrDrEs && r4logpdfgaussian_kde.logpdfsA|| ;Av!vv+ &&!5/s3004x9 o%.tG -d3 LLNNDLL1 HHdllL2d|rHc \P!V4p\P!VP\P4'gRp\ V4h\ VP4pVP4pWBV^8,,W"^8&\ \P!V44\ V48wdRp\ V4hV^8W$8,p\P!V4'dRWV, RV R2p\ V4hVPV,pVPp\WpP4VR7#)aReturn a marginal KDE distribution. Parameters ---------- dimensions : int or 1-d array_like The dimensions of the multivariate distribution corresponding with the marginal variables, that is, the indices of the dimensions that are being retained. The other dimensions are marginalized out. Returns ------- marginal_kde : gaussian_kde An object representing the marginal distribution. Notes ----- .. versionadded:: 1.10.0 zaElements of `dimensions` must be integers - the indices of the marginal variables being retained.z,All elements of `dimensions` must be unique.z Dimensions z# are invalid for a distribution in z dimensions.)rr))r,r issubdtypedtypeintegerr"r+r copyuniqueanyr)rr) r1 dimensionsdimsr2r% original_dims i_invalidr r)s && r4marginalgaussian_kde.marginals *}}Z(}}TZZ44?CS/ !   $(^+AX ryy 3t9 ,ACS/ !AX$), 66)   !9 :;,,-3l_rH) rrrr.r(r@r>rr$r rrr%)NNr)__name__ __module__ __qualname____firstlineno____doc__rrrr5rF__call__rYrdrorwrrrrr/rpropertyrrrrr)r|__annotate_func____static_attributes____classdictcell__)rs@r4rr&sk\&1%>$1L&PH3j D88<0d!F0A& ! =#~@$ B B , D/-b!!wrHc\P!W4p\P!V4PpV^8XdRpW$3#V^8XdRpW$3#VR9dRpW$3#\ V RV 24h)z Calculates the output dtype and the "spec" (=C type name). This was necessary in order to deal with the fused types in the Cython routine `gaussian_kernel_estimate`. See gh-10824 for details. r'doublez long doublez has unexpected item size: ) )r, common_typeritemsizer")r>rArCrrDs&& r4r=r=s>>*5Lxx %..H1}   Q   X    . ;H:F rH)!typesrscipyrrscipy._lib._utilrnumpyrrr r r r r rrrrrrrrrr,_statsrr _multivariater__all__rr=rrHr4rsS("/K.  @ @ FrH