Changeset 404ebbd in sasmodels for sasmodels/models/polymer_micelle.py

Timestamp:

Jul 30, 2017 12:56:22 AM (7 years ago)

Author:

Paul Kienzle <pkienzle@…>

Branches:

master, core_shell_microgels, costrafo411, magnetic_model, ticket-1257-vesicle-product, ticket_1156, ticket_1265_superball, ticket_822_more_unit_tests

Children:

48462b0

Parents:

a151caa

Message:

tuned random model generation for more models

File:

• sasmodels/models/polymer_micelle.py (modified) (7 diffs)

sasmodels/models/polymer_micelle.py

@@ -16 +16 @@
 which then defines a micelle core of $radius\_core$, which is a separate parameter
 even though it could be directly determined.
-The Gaussian random coil tails, of gyration radius $rg$, are imagined uniformly 
+The Gaussian random coil tails, of gyration radius $rg$, are imagined uniformly
 distributed around the spherical core, centred at a distance $radius\_core + d\_penetration.rg$
 from the micelle centre, where $d\_penetration$ is of order unity.
@@ -27 +27 @@
 .. math::
     P(q) = N^2\beta^2_s\Phi(qR)^2+N\beta^2_cP_c(q)+2N^2\beta_s\beta_cS_{sc}s_c(q)+N(N-1)\beta_c^2S_{cc}(q)
-    
+
     \beta_s = v\_core(sld\_core - sld\_solvent)
-    
+
     \beta_c = v\_corona(sld\_corona - sld\_solvent)
 
-where $N = n\_aggreg$, and for the spherical core of radius $R$ 
+where $N = n\_aggreg$, and for the spherical core of radius $R$
 
-.. math::   
+.. math::
    \Phi(qR)= \frac{\sin(qr) - qr\cos(qr)}{(qr)^3}
 
@@ -49 +49 @@
 
 .. math::
-   
+
    S_{sc}(q)=\Phi(qR)\psi(Z)\frac{sin(q(R+d.R_g))}{q(R+d.R_g)}
-   
+
    S_{cc}(q)=\psi(Z)^2\left[\frac{sin(q(R+d.R_g))}{q(R+d.R_g)} \right ]^2
-   
+
    \psi(Z)=\frac{[1-exp^{-Z}]}{Z}
 
@@ -60 +60 @@
 
 $P(q)$ above is multiplied by $ndensity$, and a units conversion of 10^{-13}, so $scale$
-is likely 1.0 if the scattering data is in absolute units. This model has not yet been 
+is likely 1.0 if the scattering data is in absolute units. This model has not yet been
 independently validated.
 
@@ -71 +71 @@
 """
 
-from numpy import inf
+from numpy import inf, pi
 
 name = "polymer_micelle"
@@ -80 +80 @@
 to block copolymer micelles. To work well the Gaussian chains must be much
 smaller than the core, which is often not the case.  Please study the
-reference to Pedersen and full documentation carefully. 
+reference to Pedersen and full documentation carefully.
     """
 
@@ -106 +106 @@
 source = ["lib/sas_3j1x_x.c", "polymer_micelle.c"]
 
-demo = dict(scale=1, background=0,
-            ndensity=8.94,
-            v_core=62624.0,
-            v_corona=61940.0,
-            sld_solvent=6.4,
-            sld_core=0.34,
-            sld_corona=0.8,
-            radius_core=45.0,
-            rg=20.0,
-            d_penetration=1.0,
-            n_aggreg=6.0)
-
+def random():
+    import numpy as np
+    radius_core = 10**np.random.uniform(1, 3)
+    rg = radius_core * 10**np.random.uniform(-2, -0.3)
+    d_penetration = np.random.randn()*0.05 + 1
+    n_aggreg = np.random.randint(3, 30)
+    # volume of head groups is the core volume over the number of groups,
+    # with a correction for packing fraction of the head groups.
+    v_core = 4*pi/3*radius_core**3/n_aggreg * 0.68
+    # Rg^2 for gaussian coil is a^2n/6 => a^2 = 6 Rg^2/n
+    # a=2r => r = Rg sqrt(3/2n)
+    # v = 4/3 pi r^3 n => v = 4/3 pi Rg^3 (3/2n)^(3/2) n = pi Rg^3 sqrt(6/n)
+    tail_segments = np.random.randint(6, 30)
+    v_corona = pi * rg**3 * np.sqrt(6/tail_segments)
+    V = 4*pi/3*(radius_core + rg)**3
+    pars = dict(
+        background=0,
+        scale=1e7/V,
+        ndensity=8.94,
+        v_core=v_core,
+        v_corona=v_corona,
+        radius_core=radius_core,
+        rg=rg,
+        d_penetration=d_penetration,
+        n_aggreg=n_aggreg,
+    )
+    return pars
 
 tests = [