Changeset 150fb81 in sasmodels


Ignore:
Timestamp:
Jan 11, 2017 8:52:29 AM (8 years ago)
Author:
wojciech
Branches:
master, core_shell_microgels, costrafo411, magnetic_model, ticket-1257-vesicle-product, ticket_1156, ticket_1265_superball, ticket_822_more_unit_tests
Children:
b91bb7c
Parents:
7e74ed5 (diff), b7e8b94 (diff)
Note: this is a merge changeset, the changes displayed below correspond to the merge itself.
Use the (diff) links above to see all the changes relative to each parent.
Message:

Merged with master

Files:
2 added
18 edited

Legend:

Unmodified
Added
Removed

  • sasmodels/compare.py

    r8c65a33 rfe25eda  
    10631063    # Evaluate preset parameter expressions 
    10641064    context = MATH.copy() 
     1065    context['np'] = np 
    10651066    context.update(pars) 
    10661067    context.update((k,v) for k,v in presets.items() if isinstance(v, float)) 

  • sasmodels/details.py

    r6dc78e4 rccd5f01  
    230230    npars = kernel.info.parameters.npars 
    231231    nvalues = kernel.info.parameters.nvalues 
    232     scalars = [v[0][0] for v in pairs] 
     232    scalars = [(v[0] if len(v) else np.NaN) for v, w in pairs] 
    233233    values, weights = zip(*pairs[2:npars+2]) if npars else ((),()) 
    234234    length = np.array([len(w) for w in weights]) 

  • sasmodels/kernel_header.c

    r218cdbc rb7e8b94  
    148148inline double sinc(double x) { return x==0 ? 1.0 : sin(x)/x; } 
    149149 
    150 #if 0 
     150#if 1 
     151//think cos(theta) should be sin(theta) in new coords, RKH 11Jan2017 
    151152#define ORIENT_SYMMETRIC(qx, qy, theta, phi, q, sn, cn) do { \ 
    152153    SINCOS(phi*M_PI_180, sn, cn); \ 
    153154    q = sqrt(qx*qx + qy*qy); \ 
    154     cn  = (q==0. ? 1.0 : (cn*qx + sn*qy)/q * cos(theta*M_PI_180));  \ 
     155    cn  = (q==0. ? 1.0 : (cn*qx + sn*qy)/q * sin(theta*M_PI_180));  \ 
    155156    sn = sqrt(1 - cn*cn); \ 
    156157    } while (0) 

  • sasmodels/models/barbell.py

    r0d6e865 rfcb33e4  
    2020.. math:: 
    2121 
    22     I(q) = \frac{\Delta \rho^2}{V} \left<A^2(q)\right> 
     22    I(q) = \frac{\Delta \rho^2}{V} \left<A^2(q,\alpha).sin(\alpha)\right> 
    2323 
    24 where the amplitude $A(q)$ is given as 
     24where the amplitude $A(q,\alpha)$ with the rod axis at angle $\alpha$ to $q$ is given as 
    2525 
    2626.. math:: 
    2727 
    2828    A(q) =&\ \pi r^2L 
    29         \frac{\sin\left(\tfrac12 qL\cos\theta\right)} 
    30              {\tfrac12 qL\cos\theta} 
    31         \frac{2 J_1(qr\sin\theta)}{qr\sin\theta} \\ 
     29        \frac{\sin\left(\tfrac12 qL\cos\alpha\right)} 
     30             {\tfrac12 qL\cos\alpha} 
     31        \frac{2 J_1(qr\sin\alpha)}{qr\sin\alpha} \\ 
    3232        &\ + 4 \pi R^3 \int_{-h/R}^1 dt 
    33         \cos\left[ q\cos\theta 
     33        \cos\left[ q\cos\alpha 
    3434            \left(Rt + h + {\tfrac12} L\right)\right] 
    3535        \times (1-t^2) 
    36         \frac{J_1\left[qR\sin\theta \left(1-t^2\right)^{1/2}\right]} 
    37              {qR\sin\theta \left(1-t^2\right)^{1/2}} 
     36        \frac{J_1\left[qR\sin\alpha \left(1-t^2\right)^{1/2}\right]} 
     37             {qR\sin\alpha \left(1-t^2\right)^{1/2}} 
    3838 
    3939The $\left<\ldots\right>$ brackets denote an average of the structure over 
    40 all orientations. $\left<A^2(q)\right>$ is then the form factor, $P(q)$. 
     40all orientations. $\left<A^2(q,\alpha)\right>$ is then the form factor, $P(q)$. 
    4141The scale factor is equivalent to the volume fraction of cylinders, each of 
    4242volume, $V$. Contrast $\Delta\rho$ is the difference of scattering length 
    … …  
    8585* **Author:** NIST IGOR/DANSE **Date:** pre 2010 
    8686* **Last Modified by:** Paul Butler **Date:** March 20, 2016 
    87 * **Last Reviewed by:** Paul Butler **Date:** March 20, 2016 
     87* **Last Reviewed by:** Richard Heenan **Date:** January 4, 2017 
    8888""" 
    8989from numpy import inf 

  • sasmodels/models/capped_cylinder.py

    r0d6e865 rfcb33e4  
    2121.. math:: 
    2222 
    23     I(q) = \frac{\Delta \rho^2}{V} \left<A^2(q)\right> 
     23    I(q) = \frac{\Delta \rho^2}{V} \left<A^2(q,\alpha).sin(\alpha)\right> 
    2424 
    25 where the amplitude $A(q)$ is given as 
     25where the amplitude $A(q,\alpha)$ with the rod axis at angle $\alpha$ to $q$ is given as 
    2626 
    2727.. math:: 
    2828 
    2929    A(q) =&\ \pi r^2L 
    30         \frac{\sin\left(\tfrac12 qL\cos\theta\right)} 
    31             {\tfrac12 qL\cos\theta} 
    32         \frac{2 J_1(qr\sin\theta)}{qr\sin\theta} \\ 
     30        \frac{\sin\left(\tfrac12 qL\cos\alpha\right)} 
     31            {\tfrac12 qL\cos\alpha} 
     32        \frac{2 J_1(qr\sin\alpha)}{qr\sin\alpha} \\ 
    3333        &\ + 4 \pi R^3 \int_{-h/R}^1 dt 
    34         \cos\left[ q\cos\theta 
     34        \cos\left[ q\cos\alpha 
    3535            \left(Rt + h + {\tfrac12} L\right)\right] 
    3636        \times (1-t^2) 
    37         \frac{J_1\left[qR\sin\theta \left(1-t^2\right)^{1/2}\right]} 
    38              {qR\sin\theta \left(1-t^2\right)^{1/2}} 
     37        \frac{J_1\left[qR\sin\alpha \left(1-t^2\right)^{1/2}\right]} 
     38             {qR\sin\alpha \left(1-t^2\right)^{1/2}} 
    3939 
    4040The $\left<\ldots\right>$ brackets denote an average of the structure over 
    … …  
    8888* **Author:** NIST IGOR/DANSE **Date:** pre 2010 
    8989* **Last Modified by:** Paul Butler **Date:** September 30, 2016 
    90 * **Last Reviewed by:** Richard Heenan **Date:** March 19, 2016 
     90* **Last Reviewed by:** Richard Heenan **Date:** January 4, 2017 
     91 
    9192""" 
    9293from numpy import inf 

  • sasmodels/models/core_shell_bicelle.py

    ra23639a rfcb33e4  
    4141 
    4242    I(Q,\alpha) = \frac{\text{scale}}{V_t} \cdot 
    43         F(Q,\alpha)^2 + \text{background} 
     43        F(Q,\alpha)^2.sin(\alpha) + \text{background} 
    4444 
    4545where 
    … …  
    8585* **Author:** NIST IGOR/DANSE **Date:** pre 2010 
    8686* **Last Modified by:** Paul Butler **Date:** September 30, 2016 
    87 * **Last Reviewed by:** Richard Heenan **Date:** October 5, 2016 
     87* **Last Reviewed by:** Richard Heenan **Date:** January 4, 2017 
    8888""" 
    8989 

  • sasmodels/models/core_shell_cylinder.py

    r755ecc2 rfcb33e4  
    99.. math:: 
    1010 
    11     I(q,\alpha) = \frac{\text{scale}}{V_s} F^2(q) + \text{background} 
     11    I(q,\alpha) = \frac{\text{scale}}{V_s} F^2(q,\alpha).sin(\alpha) + \text{background} 
    1212 
    1313where 
    … …  
    1515.. math:: 
    1616 
    17     F(q) = &\ (\rho_c - \rho_s) V_c 
     17    F(q,\alpha) = &\ (\rho_c - \rho_s) V_c 
    1818           \frac{\sin \left( q \tfrac12 L\cos\alpha \right)} 
    1919                {q \tfrac12 L\cos\alpha} 

  • sasmodels/models/core_shell_ellipsoid.py

    r73e08ae rb7e8b94  
    162162 
    163163q = 0.1 
    164 phi = pi/6 
    165 qx = q*cos(phi) 
    166 qy = q*sin(phi) 
    167 # After redefinition of angles find new reasonable values for unit test 
    168 #tests = [ 
    169 #    # Accuracy tests based on content in test/utest_coreshellellipsoidXTmodel.py 
    170 #    [{'radius_equat_core': 200.0, 
    171 #      'x_core': 0.1, 
    172 #      'thick_shell': 50.0, 
    173 #      'x_polar_shell': 0.2, 
    174 #      'sld_core': 2.0, 
    175 #      'sld_shell': 1.0, 
    176 #      'sld_solvent': 6.3, 
    177 #      'background': 0.001, 
    178 #      'scale': 1.0, 
    179 #     }, 1.0, 0.00189402], 
     164# tests had in old coords theta=0, phi=0; new coords theta=90, phi=0 
     165qx = q*cos(pi/6.0) 
     166qy = q*sin(pi/6.0) 
     167phi = 0.0 
     168# 11Jan2017 RKH sorted tests after redefinition of angles 
     169tests = [ 
     170     # Accuracy tests based on content in test/utest_coreshellellipsoidXTmodel.py 
     171    [{'radius_equat_core': 200.0, 
     172      'x_core': 0.1, 
     173      'thick_shell': 50.0, 
     174      'x_polar_shell': 0.2, 
     175      'sld_core': 2.0, 
     176      'sld_shell': 1.0, 
     177      'sld_solvent': 6.3, 
     178      'background': 0.001, 
     179      'scale': 1.0, 
     180     }, 1.0, 0.00189402], 
    180181 
    181182    # Additional tests with larger range of parameters 
    182 #    [{'background': 0.01}, 0.1, 11.6915], 
    183  
    184 #    [{'radius_equat_core': 20.0, 
    185 #      'x_core': 200.0, 
    186 #      'thick_shell': 54.0, 
    187 #      'x_polar_shell': 3.0, 
    188 #      'sld_core': 20.0, 
    189 #      'sld_shell': 10.0, 
    190 #      'sld_solvent': 6.0, 
    191 #      'background': 0.0, 
    192 #      'scale': 1.0, 
    193 #     }, 0.01, 8688.53], 
    194  
    195 #   [{'background': 0.001}, (0.4, 0.5), 0.00690673], 
    196  
    197 #   [{'radius_equat_core': 20.0, 
    198 #      'x_core': 200.0, 
    199 #      'thick_shell': 54.0, 
    200 #      'x_polar_shell': 3.0, 
    201 #      'sld_core': 20.0, 
    202 #      'sld_shell': 10.0, 
    203 #      'sld_solvent': 6.0, 
    204 #      'background': 0.01, 
    205 #      'scale': 0.01, 
    206 #     }, (qx, qy), 0.0100002], 
    207 #    ] 
     183    [{'background': 0.01}, 0.1, 11.6915], 
     184 
     185    [{'radius_equat_core': 20.0, 
     186      'x_core': 200.0, 
     187      'thick_shell': 54.0, 
     188      'x_polar_shell': 3.0, 
     189      'sld_core': 20.0, 
     190      'sld_shell': 10.0, 
     191      'sld_solvent': 6.0, 
     192      'background': 0.0, 
     193      'scale': 1.0, 
     194     }, 0.01, 8688.53], 
     195# why does it need theta setting here, not globally above? 
     196   [{'background': 0.001, 'theta':90.0}, (0.4, 0.5), 0.00690673], 
     197 
     198   [{'radius_equat_core': 20.0, 
     199      'x_core': 200.0, 
     200      'thick_shell': 54.0, 
     201      'x_polar_shell': 3.0, 
     202      'sld_core': 20.0, 
     203      'sld_shell': 10.0, 
     204      'sld_solvent': 6.0, 
     205      'background': 0.01, 
     206      'scale': 0.01, 
     207      'theta': 90.0, 
     208     }, (qx, qy), 0.01000025], 
     209    ] 

  • sasmodels/models/cylinder.py

    r4cdd0cc rb7e8b94  
    22# Note: model title and parameter table are inserted automatically 
    33r""" 
    4 The form factor is normalized by the particle volume V = \piR^2L. 
     4 
    55For information about polarised and magnetic scattering, see 
    66the :ref:`magnetism` documentation. 
    … …  
    1414.. math:: 
    1515 
    16     P(q,\alpha) = \frac{\text{scale}}{V} F^2(q,\alpha) + \text{background} 
     16    P(q,\alpha) = \frac{\text{scale}}{V} F^2(q,\alpha).sin(\alpha) + \text{background} 
    1717 
    1818where 
    … …  
    2525           \frac{J_1 \left(q R \sin \alpha\right)}{q R \sin \alpha} 
    2626 
    27 and $\alpha$ is the angle between the axis of the cylinder and $\vec q$, $V$ 
     27and $\alpha$ is the angle between the axis of the cylinder and $\vec q$, $V =\pi R^2L$ 
    2828is the volume of the cylinder, $L$ is the length of the cylinder, $R$ is the 
    2929radius of the cylinder, and $\Delta\rho$ (contrast) is the scattering length 
    … …  
    3535.. math:: 
    3636 
    37     F^2(q)=\int_{0}^{\pi/2}{F^2(q,\theta)\sin(\theta)d\theta} 
     37    F^2(q)=\int_{0}^{\pi/2}{F^2(q,\alpha)\sin(\alpha)d\alpha}=\int_{0}^{1}{F^2(q,u)du} 
    3838 
    3939 
    40 To provide easy access to the orientation of the cylinder, we define the 
    41 axis of the cylinder using two angles $\theta$ and $\phi$. Those angles 
     40Numerical integration is simplified by a change of variable to $u = cos(\alpha)$ with  
     41$sin(\alpha)=\sqrt{1-u^2}$.  
     42 
     43The output of the 1D scattering intensity function for randomly oriented 
     44cylinders is thus given by 
     45 
     46.. math:: 
     47 
     48    P(q) = \frac{\text{scale}}{V} 
     49        \int_0^{\pi/2} F^2(q,\alpha) \sin \alpha\ d\alpha + \text{background} 
     50 
     51 
     52NB: The 2nd virial coefficient of the cylinder is calculated based on the 
     53radius and length values, and used as the effective radius for $S(q)$ 
     54when $P(q) \cdot S(q)$ is applied. 
     55 
     56For oriented cylinders, we define the direction of the 
     57axis of the cylinder using two angles $\theta$ (note this is not the 
     58same as the scattering angle used in q) and $\phi$. Those angles 
    4259are defined in :numref:`cylinder-angle-definition` . 
    4360 
    … …  
    4865    Definition of the angles for oriented cylinders. 
    4966 
    50  
    51 NB: The 2nd virial coefficient of the cylinder is calculated based on the 
    52 radius and length values, and used as the effective radius for $S(q)$ 
    53 when $P(q) \cdot S(q)$ is applied. 
    54  
    55 The output of the 1D scattering intensity function for randomly oriented 
    56 cylinders is then given by 
    57  
    58 .. math:: 
    59  
    60     P(q) = \frac{\text{scale}}{V} 
    61         \int_0^{\pi/2} F^2(q,\alpha) \sin \alpha\ d\alpha + \text{background} 
    62  
    63 The $\theta$ and $\phi$ parameters are not used for the 1D output. 
     67The $\theta$ and $\phi$ parameters only appear in the model when fitting 2d data. 
    6468 
    6569Validation 
    … …  
    7478 
    7579    P(q) = \int_0^{\pi/2} d\phi 
    76         \int_0^\pi p(\alpha) P_0(q,\alpha) \sin \alpha\ d\alpha 
     80        \int_0^\pi p(\theta) P_0(q,\theta) \sin \theta\ d\theta 
    7781 
    7882 
    7983where $p(\theta,\phi) = 1$ is the probability distribution for the orientation 
    80 and $P_0(q,\alpha)$ is the scattering intensity for the fully oriented 
     84and $P_0(q,\theta)$ is the scattering intensity for the fully oriented 
    8185system, and then comparing to the 1D result. 
    8286 
    … …  
    145149 
    146150qx, qy = 0.2 * np.cos(2.5), 0.2 * np.sin(2.5) 
    147 # After redefinition of angles, find new tests values  
    148 #tests = [[{}, 0.2, 0.042761386790780453], 
    149 #         [{}, [0.2], [0.042761386790780453]], 
    150 #         [{'theta':10.0, 'phi':10.0}, (qx, qy), 0.03514647218513852], 
    151 #         [{'theta':10.0, 'phi':10.0}, [(qx, qy)], [0.03514647218513852]], 
    152 #        ] 
     151# After redefinition of angles, find new tests values.  Was 10 10 in old coords 
     152tests = [[{}, 0.2, 0.042761386790780453], 
     153        [{}, [0.2], [0.042761386790780453]], 
     154#  new coords     
     155        [{'theta':80.1534480601659, 'phi':10.1510817110481}, (qx, qy), 0.03514647218513852], 
     156        [{'theta':80.1534480601659, 'phi':10.1510817110481}, [(qx, qy)], [0.03514647218513852]], 
     157# old coords   [{'theta':10.0, 'phi':10.0}, (qx, qy), 0.03514647218513852], 
     158#              [{'theta':10.0, 'phi':10.0}, [(qx, qy)], [0.03514647218513852]], 
     159        ] 
    153160del qx, qy  # not necessary to delete, but cleaner 
    154161# ADDED by:  RKH  ON: 18Mar2016 renamed sld's etc 

  • sasmodels/models/elliptical_cylinder.py

    ra807206 rfcb33e4  
    2020 
    2121    I(\vec q)=\frac{1}{V_\text{cyl}}\int{d\psi}\int{d\phi}\int{ 
    22         p(\theta,\phi,\psi)F^2(\vec q,\alpha,\psi)\sin(\theta)d\theta} 
     22        p(\theta,\phi,\psi)F^2(\vec q,\alpha,\psi)\sin(\alpha)d\alpha} 
    2323 
    2424with the functions 
    … …  
    2626.. math:: 
    2727 
    28     F(\vec q,\alpha,\psi) = 2\frac{J_1(a)\sin(b)}{ab} 
     28    F(q,\alpha,\psi) = 2\frac{J_1(a)\sin(b)}{ab} 
    2929 
    3030where 
    … …  
    3232.. math:: 
    3333 
    34     a &= \vec q\sin(\alpha)\left[ 
    35         r^2_\text{major}\sin^2(\psi)+r^2_\text{minor}\cos(\psi) \right]^{1/2} 
     34    a = qr'\sin(\alpha) 
     35     
     36    b = q\frac{L}{2}\cos(\alpha) 
     37     
     38    r'=\frac{r_{minor}}{\sqrt{2}}\sqrt{(1+\nu^{2}) + (1-\nu^{2})cos(\psi)} 
    3639 
    37     b &= \vec q\frac{L}{2}\cos(\alpha) 
    3840 
    39 and the angle $\Psi$ is defined as the orientation of the major axis of the 
     41and the angle $\psi$ is defined as the orientation of the major axis of the 
    4042ellipse with respect to the vector $\vec q$. The angle $\alpha$ is the angle 
    4143between the axis of the cylinder and $\vec q$. 
    … …  
    9597 
    9698L A Feigin and D I Svergun, *Structure Analysis by Small-Angle X-Ray and 
    97 Neutron Scattering*, Plenum, New York, (1987) 
     99Neutron Scattering*, Plenum, New York, (1987) [see table 3.4] 
     100 
     101Authorship and Verification 
     102---------------------------- 
     103 
     104* **Author:** 
     105* **Last Modified by:**  
     106* **Last Reviewed by:**  Richard Heenan - corrected equation in docs **Date:** December 21, 2016 
     107 
    98108""" 
    99109 

  • sasmodels/models/lib/polevl.c

    r0278e3f r447e9aa  
    2828 *            N                   0 
    2929 * 
    30  *  The function p1evl() assumes that coef[N] = 1.0 and is 
     30 * The function p1evl() assumes that C_N = 1.0 and is 
    3131 * omitted from the array.  Its calling arguments are 
    3232 * otherwise the same as polevl(). 

  • sasmodels/models/lib/sas_J0.c

    r1596de3 rc8902ac  
    5555#if FLOAT_SIZE>4 
    5656//Cephes double precission 
    57 double j0(double x); 
     57double cephes_j0(double x); 
    5858 
    5959 constant double PPJ0[8] = { 
    … …  
    147147  }; 
    148148 
    149 double j0(double x) 
     149double cephes_j0(double x) 
    150150{ 
    151151    double w, z, p, q, xn; 
    … …  
    186186#else 
    187187//Cephes single precission 
    188 float j0f(float x); 
     188float cephes_j0f(float x); 
    189189 
    190190 constant float MOJ0[8] = { 
    … …  
    221221 }; 
    222222 
    223 float j0f(float x) 
     223float cephes_j0f(float x) 
    224224{ 
    225225    float xx, w, z, p, q, xn; 
    … …  
    257257 
    258258#if FLOAT_SIZE>4 
    259 #define sas_J0 j0 
     259#define sas_J0 cephes_j0 
    260260#else 
    261 #define sas_J0 j0f 
     261#define sas_J0 cephes_j0f 
    262262#endif 

  • sasmodels/models/lib/sas_J1.c

    r1596de3 rc8902ac  
    4242#if FLOAT_SIZE>4 
    4343//Cephes double pression function 
    44 double j1(double x); 
     44double cephes_j1(double x); 
    4545 
    4646constant double RPJ1[8] = { 
    … …  
    106106    0.0 }; 
    107107 
    108 double j1(double x) 
     108double cephes_j1(double x) 
    109109{ 
    110110 
    … …  
    144144#else 
    145145//Single precission version of cephes 
    146 float j1f(float x); 
     146float cephes_j1f(float x); 
    147147 
    148148constant float JPJ1[8] = { 
    … …  
    179179    }; 
    180180 
    181 float j1f(float x) 
     181float cephes_j1f(float x) 
    182182{ 
    183183 
    … …  
    211211 
    212212#if FLOAT_SIZE>4 
    213 #define sas_J1 j1 
     213#define sas_J1 cephes_j1 
    214214#else 
    215 #define sas_J1 j1f 
     215#define sas_J1 cephes_j1f 
    216216#endif 
    217217 

  • sasmodels/models/lib/sas_JN.c

    r1596de3 rc8902ac  
    5050#if FLOAT_SIZE > 4 
    5151 
    52 double jn( int n, double x ); 
    53 double jn( int n, double x ) { 
     52double cephes_jn( int n, double x ); 
     53double cephes_jn( int n, double x ) { 
    5454 
    5555    // PAK: seems to be machine epsilon/2 
    … …  
    7575 
    7676    if( n == 0 ) 
    77         return( sign * j0(x) ); 
     77        return( sign * cephes_j0(x) ); 
    7878    if( n == 1 ) 
    79         return( sign * j1(x) ); 
     79        return( sign * cephes_j1(x) ); 
    8080    if( n == 2 ) 
    81         return( sign * (2.0 * j1(x) / x  -  j0(x)) ); 
     81        return( sign * (2.0 * cephes_j1(x) / x  -  cephes_j0(x)) ); 
    8282 
    8383    if( x < MACHEP ) 
    … …  
    112112 
    113113    if( fabs(pk) > fabs(pkm1) ) 
    114         ans = j1(x)/pk; 
     114        ans = cephes_j1(x)/pk; 
    115115    else 
    116         ans = j0(x)/pkm1; 
     116        ans = cephes_j0(x)/pkm1; 
    117117 
    118118    return( sign * ans ); 
    … …  
    121121#else 
    122122 
    123 float jnf(int n, float x); 
    124 float jnf(int n, float x) 
     123float cephes_jnf(int n, float x); 
     124float cephes_jnf(int n, float x) 
    125125{ 
    126126    // PAK: seems to be machine epsilon/2 
    … …  
    146146 
    147147    if( n == 0 ) 
    148         return( sign * j0f(x) ); 
     148        return( sign * cephes_j0f(x) ); 
    149149    if( n == 1 ) 
    150         return( sign * j1f(x) ); 
     150        return( sign * cephes_j1f(x) ); 
    151151    if( n == 2 ) 
    152         return( sign * (2.0 * j1f(x) / x  -  j0f(x)) ); 
     152        return( sign * (2.0 * cephes_j1f(x) / x  -  cephes_j0f(x)) ); 
    153153 
    154154    if( x < MACHEP ) 
    … …  
    189189 
    190190    if( r > ans )  /* if( fabs(pk) > fabs(pkm1) ) */ 
    191         ans = sign * j1f(x)/pk; 
     191        ans = sign * cephes_j1f(x)/pk; 
    192192    else 
    193         ans = sign * j0f(x)/pkm1; 
     193        ans = sign * cephes_j0f(x)/pkm1; 
    194194    return( ans ); 
    195195} 
    … …  
    197197 
    198198#if FLOAT_SIZE>4 
    199 #define sas_JN jn 
     199#define sas_JN cephes_jn 
    200200#else 
    201 #define sas_JN jnf 
     201#define sas_JN cephes_jnf 
    202202#endif 

  • sasmodels/models/stacked_disks.py

    r07300ea rb7e8b94  
    148148            sld_layer=0.0, 
    149149            sld_solvent=5.0, 
    150             theta=0, 
     150            theta=90, 
    151151            phi=0) 
    152 #After redefinition to spherical coordinates find new reasonable test values 
    153 #tests = [ 
    154 #    # Accuracy tests based on content in test/utest_extra_models.py. 
    155 #    # Added 2 tests with n_stacked = 5 using SasView 3.1.2 - PDB 
    156 #    [{'thick_core': 10.0, 
    157 #      'thick_layer': 15.0, 
    158 #      'radius': 3000.0, 
    159 #      'n_stacking': 1.0, 
    160 #      'sigma_d': 0.0, 
    161 #      'sld_core': 4.0, 
    162 #      'sld_layer': -0.4, 
    163 #      'solvent_sd': 5.0, 
    164 #      'theta': 0.0, 
    165 #      'phi': 0.0, 
    166 #      'scale': 0.01, 
    167 #      'background': 0.001, 
    168 #     }, 0.001, 5075.12], 
    169  
    170 #    [{'thick_core': 10.0, 
    171 #      'thick_layer': 15.0, 
    172 #      'radius': 3000.0, 
    173 #      'n_stacking': 5.0, 
    174 #      'sigma_d': 0.0, 
    175 #      'sld_core': 4.0, 
    176 #      'sld_layer': -0.4, 
    177 #      'solvent_sd': 5.0, 
    178 #      'theta': 0.0, 
    179 #      'phi': 0.0, 
    180 #      'scale': 0.01, 
    181 #      'background': 0.001, 
    182 #     }, 0.001, 5065.12793824], 
    183  
    184 #    [{'thick_core': 10.0, 
    185 #      'thick_layer': 15.0, 
    186 #      'radius': 3000.0, 
    187 #      'n_stacking': 5.0, 
    188 #      'sigma_d': 0.0, 
    189 #      'sld_core': 4.0, 
    190 #      'sld_layer': -0.4, 
    191 #      'solvent_sd': 5.0, 
    192 #      'theta': 0.0, 
    193 #      'phi': 0.0, 
    194 #      'scale': 0.01, 
    195 #      'background': 0.001, 
    196 #     }, 0.164, 0.0127673597265], 
    197  
    198 #    [{'thick_core': 10.0, 
    199 #      'thick_layer': 15.0, 
    200 #      'radius': 3000.0, 
    201 #      'n_stacking': 1.0, 
    202 #      'sigma_d': 0.0, 
    203 #      'sld_core': 4.0, 
    204 #      'sld_layer': -0.4, 
    205 #      'solvent_sd': 5.0, 
    206 #      'theta': 0.0, 
    207 #      'phi': 0.0, 
    208 #      'scale': 0.01, 
    209 #      'background': 0.001, 
    210 #     }, [0.001, 90.0], [5075.12, 0.001]], 
    211  
    212 #    [{'thick_core': 10.0, 
    213 #      'thick_layer': 15.0, 
    214 #      'radius': 3000.0, 
    215 #      'n_stacking': 1.0, 
    216 #      'sigma_d': 0.0, 
    217 #      'sld_core': 4.0, 
    218 #      'sld_layer': -0.4, 
    219 #      'solvent_sd': 5.0, 
    220 #      'theta': 0.0, 
    221 #      'phi': 0.0, 
    222 #      'scale': 0.01, 
    223 #      'background': 0.001, 
    224 #     }, ([0.4, 0.5]), [0.00105074, 0.00121761]], 
    225  
    226 #    [{'thick_core': 10.0, 
    227 #      'thick_layer': 15.0, 
    228 #      'radius': 3000.0, 
    229 #      'n_stacking': 1.0, 
    230 #      'sigma_d': 0.0, 
    231 #      'sld_core': 4.0, 
    232 #      'sld_layer': -0.4, 
    233 #     'solvent_sd': 5.0, 
    234 #      'theta': 0.0, 
    235 #      'phi': 0.0, 
    236 #      'scale': 0.01, 
    237 #      'background': 0.001, 
    238 #     }, ([1.3, 1.57]), [0.0010039, 0.0010038]], 
    239 #    ] 
    240 # 21Mar2016   RKH notes that unit tests all have n_stacking=1, ought to test other values 
    241  
     152# After redefinition of spherical coordinates - 
     153# tests had in old coords theta=0, phi=0; new coords theta=90, phi=0 
     154# but should not matter here as so far all the tests are 1D not 2D 
     155tests = [ 
     156# Accuracy tests based on content in test/utest_extra_models.py. 
     157# Added 2 tests with n_stacked = 5 using SasView 3.1.2 - PDB; for which alas q=0.001 values seem closer to n_stacked=1 not 5, changed assuming my 4.1 code OK, RKH 
     158    [{'thick_core': 10.0, 
     159      'thick_layer': 15.0, 
     160      'radius': 3000.0, 
     161      'n_stacking': 1.0, 
     162      'sigma_d': 0.0, 
     163      'sld_core': 4.0, 
     164      'sld_layer': -0.4, 
     165      'solvent_sd': 5.0, 
     166      'theta': 90.0, 
     167      'phi': 0.0, 
     168      'scale': 0.01, 
     169      'background': 0.001, 
     170     }, 0.001, 5075.12], 
     171    [{'thick_core': 10.0, 
     172      'thick_layer': 15.0, 
     173      'radius': 3000.0, 
     174      'n_stacking': 5, 
     175      'sigma_d': 0.0, 
     176      'sld_core': 4.0, 
     177      'sld_layer': -0.4, 
     178      'solvent_sd': 5.0, 
     179      'theta': 90.0, 
     180      'phi': 0.0, 
     181      'scale': 0.01, 
     182      'background': 0.001, 
     183#     }, 0.001, 5065.12793824],    n_stacking=1 not 5 ? slight change in value here 11jan2017, check other cpu types 
     184#     }, 0.001, 5075.11570493], 
     185     }, 0.001, 25325.635693], 
     186    [{'thick_core': 10.0, 
     187      'thick_layer': 15.0, 
     188      'radius': 3000.0, 
     189      'n_stacking': 5, 
     190      'sigma_d': 0.0, 
     191      'sld_core': 4.0, 
     192      'sld_layer': -0.4, 
     193      'solvent_sd': 5.0, 
     194      'theta': 90.0, 
     195      'phi': 0.0, 
     196      'scale': 0.01, 
     197      'background': 0.001, 
     198#     }, 0.164, 0.0127673597265],    n_stacking=1 not 5 ?  slight change in value here 11jan2017, check other cpu types 
     199#     }, 0.164, 0.01480812968], 
     200     }, 0.164, 0.0598367986509], 
     201 
     202    [{'thick_core': 10.0, 
     203      'thick_layer': 15.0, 
     204      'radius': 3000.0, 
     205      'n_stacking': 1.0, 
     206      'sigma_d': 0.0, 
     207      'sld_core': 4.0, 
     208      'sld_layer': -0.4, 
     209      'solvent_sd': 5.0, 
     210      'theta': 90.0, 
     211      'phi': 0.0, 
     212      'scale': 0.01, 
     213      'background': 0.001, 
     214# second test here was at q=90, changed it to q=5, note I(q) is then just value of flat background 
     215     }, [0.001, 5.0], [5075.12, 0.001]], 
     216 
     217    [{'thick_core': 10.0, 
     218      'thick_layer': 15.0, 
     219      'radius': 3000.0, 
     220      'n_stacking': 1.0, 
     221      'sigma_d': 0.0, 
     222      'sld_core': 4.0, 
     223      'sld_layer': -0.4, 
     224      'solvent_sd': 5.0, 
     225      'theta': 90.0, 
     226      'phi': 0.0, 
     227      'scale': 0.01, 
     228      'background': 0.001, 
     229     }, ([0.4, 0.5]), [0.00105074, 0.00121761]], 
     230 
     231    [{'thick_core': 10.0, 
     232      'thick_layer': 15.0, 
     233      'radius': 3000.0, 
     234      'n_stacking': 1.0, 
     235      'sigma_d': 0.0, 
     236      'sld_core': 4.0, 
     237      'sld_layer': -0.4, 
     238     'solvent_sd': 5.0, 
     239      'theta': 90.0, 
     240      'phi': 0.0, 
     241      'scale': 0.01, 
     242      'background': 0.001, 
     243     }, ([1.3, 1.57]), [0.0010039, 0.0010038]], 
     244    ] 
     245# 11Jan2017   RKH checking unit test agai, note they are all 1D, no 2D 
     246 

  • sasmodels/resolution2d.py

    r40a87fa r7e94989  
    6464        # just need q_calc and weights 
    6565        self.data = data 
    66         self.index = index 
    67  
    68         self.qx_data = data.qx_data[index] 
    69         self.qy_data = data.qy_data[index] 
    70         self.q_data = data.q_data[index] 
     66        self.index = index if index is not None else slice(None) 
     67 
     68        self.qx_data = data.qx_data[self.index] 
     69        self.qy_data = data.qy_data[self.index] 
     70        self.q_data = data.q_data[self.index] 
    7171 
    7272        dqx = getattr(data, 'dqx_data', None) 
    … …  
    7474        if dqx is not None and dqy is not None: 
    7575            # Here dqx and dqy mean dq_parr and dq_perp 
    76             self.dqx_data = dqx[index] 
    77             self.dqy_data = dqy[index] 
     76            self.dqx_data = dqx[self.index] 
     77            self.dqy_data = dqy[self.index] 
    7878            ## Remove singular points if exists 
    7979            self.dqx_data[self.dqx_data < SIGMA_ZERO] = SIGMA_ZERO 
    … …  
    126126        # The angle (phi) of the original q point 
    127127        q_phi = np.arctan(q_phi).repeat(nbins)\ 
    128             .reshape(nq, nbins).transpose().flatten() 
     128            .reshape([nq, nbins]).transpose().flatten() 
    129129        ## Find Gaussian weight for each dq bins: The weight depends only 
    130130        #  on r-direction (The integration may not need) 

  • sasmodels/sasview_model.py

    r8977226 r64614ad  
    583583                            % type(qdist)) 
    584584 
     585    def get_composition_models(self): 
     586        """ 
     587            Returns usable models that compose this model 
     588        """ 
     589        s_model = None 
     590        p_model = None 
     591        if hasattr(self._model_info, "composition") \ 
     592           and self._model_info.composition is not None: 
     593            p_model = _make_model_from_info(self._model_info.composition[1][0])() 
     594            s_model = _make_model_from_info(self._model_info.composition[1][1])() 
     595        return p_model, s_model 
     596 
    585597    def calculate_Iq(self, qx, qy=None): 
    586598        # type: (Sequence[float], Optional[Sequence[float]]) -> np.ndarray 

  • doc/ref/gpu/gpu_computations.rst

    r3f5a566 r7e74ed5  
    3131from available OpenCL platforms. 
    3232 
     33OpenCL devices can be set from OpenCL options dialog in Fitting menu or as 
     34enviromental variables. 
     35 
     36**If you don't want to use OpenCL, you can select "No OpenCL" option from** 
     37**GUI dialog or set *SAS_OPENCL=None* in your environment settings** 
     38**This will only use normal programs.** 
     39 
    3340SasView prefers AMD or NVIDIA drivers for GPU, and prefers Intel or 
    3441Apple drivers for CPU. Both GPU and CPU are included on the assumption that CPU  
    … …  
    3946chose to run the model. 
    4047 
    41 **If you don't want to use OpenCL, you can set** *SAS_OPENCL=None* 
    42 **in your environment settings, and it will only use normal programs.** 
    43  
    44 If you want to use one of the other devices, you can run the following 
    45 from the python console in SasView:: 
     48If you want to use one of the other devices, you can select it from OpenCL 
     49options dialog (accessible from Fitting menu) or run the following from 
     50the python console in SasView:: 
    4651 
    4752    import pyopencl as cl 
Note: See TracChangeset for help on using the changeset viewer.