def __init__(self,
hooft_analysis,
use_students_t_distribution=False,
students_t_nu=None,
probability_plot_slope=None):
self.delta_fo2, minus_fo2 =\
hooft_analysis.delta_fo2.generate_bijvoet_mates().hemispheres_acentrics()
self.delta_fc2, minus_fc2 =\
hooft_analysis.delta_fc2.generate_bijvoet_mates().hemispheres_acentrics()
# we want to plot both hemispheres
self.delta_fo2.indices().extend(minus_fo2.indices())
self.delta_fo2.data().extend(minus_fo2.data() * -1)
self.delta_fo2.sigmas().extend(minus_fo2.sigmas())
self.delta_fc2.indices().extend(minus_fc2.indices())
self.delta_fc2.data().extend(minus_fc2.data() * -1)
self.indices = self.delta_fo2.indices()
observed_deviations = (hooft_analysis.G * self.delta_fc2.data()
- self.delta_fo2.data())/self.delta_fo2.sigmas()
if probability_plot_slope is not None:
observed_deviations /= probability_plot_slope
selection = flex.sort_permutation(observed_deviations)
observed_deviations = observed_deviations.select(selection)
if use_students_t_distribution:
if students_t_nu is None:
students_t_nu = maximise_students_t_correlation_coefficient(
observed_deviations, 1, 200)
self.distribution = distributions.students_t_distribution(students_t_nu)
else:
self.distribution = distributions.normal_distribution()
self.x = self.distribution.quantiles(observed_deviations.size())
self.y = observed_deviations
self.fit = flex.linear_regression(self.x[5:-5], self.y[5:-5])
self.correlation = flex.linear_correlation(self.x[5:-5], self.y[5:-5])
assert self.fit.is_well_defined()
def absolute_structure_analysis(xs, fo2, fc, scale, nu=None, log=None,
outlier_cutoff_factor=None):
if log is None:
log = sys.stdout
hooft_analysis = absolute_structure.hooft_analysis(
fo2, fc, scale_factor=scale, outlier_cutoff_factor=outlier_cutoff_factor)
print >> log, "Gaussian analysis:"
hooft_analysis.show(out=log)
NPP = absolute_structure.bijvoet_differences_probability_plot(
hooft_analysis)
print >> log, "Probability plot:"
NPP.show(out=log)
print >> log
if nu is None:
nu = absolute_structure.maximise_students_t_correlation_coefficient(
NPP.y, min_nu=1, max_nu=200)
distribution = distributions.students_t_distribution(nu)
observed_deviations = NPP.y
expected_deviations = distribution.quantiles(observed_deviations.size())
fit = flex.linear_regression(
expected_deviations[5:-5], observed_deviations[5:-5])
t_analysis = absolute_structure.students_t_hooft_analysis(
fo2, fc, nu, scale_factor=scale, probability_plot_slope=fit.slope(),
outlier_cutoff_factor=outlier_cutoff_factor)
tPP = absolute_structure.bijvoet_differences_probability_plot(
t_analysis, use_students_t_distribution=True, students_t_nu=nu)
print >> log, "Student's t analysis:"
print >> log, "nu: %.2f" %nu
t_analysis.show(out=log)
print >> log, "Probability plot:"
tPP.show(out=log)
print >> log
if xs is not None:
flack = absolute_structure.flack_analysis(xs, fo2.as_xray_observations())
flack.show(out=log)
def __init__(self,
hooft_analysis,
use_students_t_distribution=False,
students_t_nu=None,
probability_plot_slope=None):
self.delta_fo2, minus_fo2 =\
hooft_analysis.delta_fo2.generate_bijvoet_mates().hemispheres_acentrics()
self.delta_fc2, minus_fc2 =\
hooft_analysis.delta_fc2.generate_bijvoet_mates().hemispheres_acentrics()
# we want to plot both hemispheres
self.delta_fo2.indices().extend(minus_fo2.indices())
self.delta_fo2.data().extend(minus_fo2.data() * -1)
self.delta_fo2.sigmas().extend(minus_fo2.sigmas())
self.delta_fc2.indices().extend(minus_fc2.indices())
self.delta_fc2.data().extend(minus_fc2.data() * -1)
self.indices = self.delta_fo2.indices()
observed_deviations = (hooft_analysis.G * self.delta_fc2.data()
- self.delta_fo2.data())/self.delta_fo2.sigmas()
if probability_plot_slope is not None:
observed_deviations /= probability_plot_slope
selection = flex.sort_permutation(observed_deviations)
observed_deviations = observed_deviations.select(selection)
if use_students_t_distribution:
if students_t_nu is None:
students_t_nu = maximise_students_t_correlation_coefficient(
observed_deviations, 1, 200)
self.distribution = distributions.students_t_distribution(students_t_nu)
else:
self.distribution = distributions.normal_distribution()
self.x = self.distribution.quantiles(observed_deviations.size())
self.y = observed_deviations
self.fit = flex.linear_regression(self.x[5:-5], self.y[5:-5])
self.correlation = flex.linear_correlation(self.x[5:-5], self.y[5:-5])
assert self.fit.is_well_defined()
def compute_cc_significance(r, n, p):
# https://en.wikipedia.org/wiki/Pearson_product-moment_correlation_coefficient#Testing_using_Student.27s_t-distribution
if r == -1 or n <= 2:
significance = False
critical_value = 0
else:
from scitbx.math import distributions
dist = distributions.students_t_distribution(n - 2)
t = dist.quantile(1 - p)
critical_value = t / sqrt(n - 2 + t**2)
significance = r > critical_value
return significance, critical_value
def absolute_structure_analysis(xs,
fo2,
fc,
scale,
nu=None,
log=None,
outlier_cutoff_factor=None):
if log is None:
log = sys.stdout
hooft_analysis = absolute_structure.hooft_analysis(
fo2,
fc,
scale_factor=scale,
outlier_cutoff_factor=outlier_cutoff_factor)
print >> log, "Gaussian analysis:"
hooft_analysis.show(out=log)
NPP = absolute_structure.bijvoet_differences_probability_plot(
hooft_analysis)
print >> log, "Probability plot:"
NPP.show(out=log)
print >> log
if nu is None:
nu = absolute_structure.maximise_students_t_correlation_coefficient(
NPP.y, min_nu=1, max_nu=200)
distribution = distributions.students_t_distribution(nu)
observed_deviations = NPP.y
expected_deviations = distribution.quantiles(observed_deviations.size())
fit = flex.linear_regression(expected_deviations[5:-5],
observed_deviations[5:-5])
t_analysis = absolute_structure.students_t_hooft_analysis(
fo2,
fc,
nu,
scale_factor=scale,
probability_plot_slope=fit.slope(),
outlier_cutoff_factor=outlier_cutoff_factor)
tPP = absolute_structure.bijvoet_differences_probability_plot(
t_analysis, use_students_t_distribution=True, students_t_nu=nu)
print >> log, "Student's t analysis:"
print >> log, "nu: %.2f" % nu
t_analysis.show(out=log)
print >> log, "Probability plot:"
tPP.show(out=log)
print >> log
if xs is not None:
flack = absolute_structure.flack_analysis(xs,
fo2.as_xray_observations())
flack.show(out=log)
from cctbx.array_family import flex
from iotbx import csv_utils
import libtbx
import libtbx.utils
from libtbx.test_utils import approx_equal
import scitbx.random
from scitbx.random import variate, normal_distribution, gamma_distribution
from scitbx.math import distributions
from smtbx import absolute_structure
try:
distributions.students_t_distribution(1)
except RuntimeError as e:
# XXX Student's t distribution is not supported with GCC 3.2 builds
if str(e).startswith("Implementation not available in this build."):
students_t_available = False
print("Skipping exercise_hooft_analysis() with Student's t distribution.")
else:
raise RuntimeError(e)
else:
students_t_available = True
class test_case(object):
d_min=1
use_students_t_errors=False
elements = ("N", "C", "C", "S") * 5
from cctbx.array_family import flex
from iotbx import csv_utils
import libtbx
import libtbx.utils
from libtbx.test_utils import approx_equal
import scitbx.random
from scitbx.random import variate, normal_distribution, gamma_distribution
from scitbx.math import distributions
from smtbx import absolute_structure
try:
distributions.students_t_distribution(1)
except RuntimeError, e:
# XXX Student's t distribution is not supported with GCC 3.2 builds
if str(e).startswith("Implementation not available in this build."):
students_t_available = False
print "Skipping exercise_hooft_analysis() with Student's t distribution."
else:
raise RuntimeError(e)
else:
students_t_available = True
class test_case(object):
d_min=1
use_students_t_errors=False
elements = ("N", "C", "C", "S") * 5
def compute_corr_coeff(i):
distribution = distributions.students_t_distribution(i)
expected_deviations = distribution.quantiles(observed_deviations.size())
return flex.linear_correlation(
observed_deviations[5:-5], expected_deviations[5:-5])
def compute_corr_coeff(i):
distribution = distributions.students_t_distribution(i)
expected_deviations = distribution.quantiles(observed_deviations.size())
return flex.linear_correlation(
observed_deviations[5:-5], expected_deviations[5:-5])
