Compare reflecitivity results from different calculations

1- Classic 1D specular reflectivity

2- Constant-Q summing for 1D specular reflectivity

3- Summing the specular ridge after computing the off-specular in (kf_z - kf_i) vs Qz with simple rebinning. This approach just computes (kf_z - kf_i) and Qz for each pixel and TOF bin and histograms them.

4- The "smoothing" approach, which is similar to 3, but applies Gaussian smoothing on top. This was the approach followed in QuickNXS 1.0. It produces a smooth image of the off-specular signal but doesn't provide errors for data analysis and can't be modeled directly.

The different calculations are compared for a TOF binning of 40 usec and 400 usec to see the effect of TOF binning on the final reflectivity curve.

All calculations use SANGLE. The TOF distribution has been trimmed to match a 2.8 A bandwidth, which cuts the tails on each side of the TOF distrubution (a small effect, but makes the data cleaner).



In [2]:

    
import os
import numpy as np
import scipy.stats
import plotly.offline as py
import plotly.graph_objs as go
py.init_notebook_mode(connected=True)

import warnings
warnings.filterwarnings('ignore', module='numpy')
warnings.filterwarnings('ignore')



In [3]:

    
import sys
sys.path.append("/SNS/users/m2d/git/reflectivity_ui/test/notebooks")
from plot_utils import plot1d, plot_heatmap



In [4]:

    
def load_data(file_path, index=75, q_axis=True, width=0):
    offspec_data = np.loadtxt(file_path)
    print offspec_data.shape

    unique_x = np.unique(offspec_data.T[0])
    xbins = len(unique_x)
    unique_x = np.unique(offspec_data.T[1])
    zbins = len(unique_x)

    print("\n\nFile: %s" % file_path)
    print("Binning: %s %s [%s]" %(xbins, zbins, xbins*zbins))

    off = np.reshape(offspec_data, (zbins, xbins, offspec_data.shape[1]))

    dk = (off.T[0]).T[0]
    qz = (off.T[1])[0]

    quick_intensity = off.T[2].T
    
    if q_axis:
        if width>0:
            refl = np.sum(quick_intensity.T[index-width:index+width+1], axis=0)
            refl /= 1.0 #width
        else:
            refl = quick_intensity.T[index]
        _index_value = dk[index]
    else:
        refl = quick_intensity[index]
        _index_value = qz[index]

    scale = np.max(quick_intensity)
    #scale = np.max(refl)
    #scale = 1
    quick_intensity /= scale
    quick_intensity[quick_intensity<1.0e-9] = 1.0e-9

    print("Scale = %s" % scale)
    print("Const value = %s" % _index_value)
    print("Width = %s" % width)
    return [qz, refl/scale], dk, qz, quick_intensity

Smooth dTOF=400 usec



In [6]:

    
_spec_400, dk, qz, off_spec = load_data('/SNS/users/m2d/git/reflectivity_ui/test/notebooks/data/REF_M_30889+30890_OffSpecSmooth_Off_Off_v2_TOF_400.dat',
                                       index=42, q_axis=True, width=5)

plot_heatmap(dk, qz, np.log(off_spec), x_title='', y_title='', surface=False,
                 x_log=False, y_log=False)









    



(23688, 3)


File: /SNS/users/m2d/git/reflectivity_ui/test/notebooks/data/REF_M_30889+30890_OffSpecSmooth_Off_Off_v2_TOF_400.dat
Binning: 84 282 [23688]
Scale = 9.30532
Const value = 0.000180723
Width = 5

Binned dTOF=400 usec



In [7]:

    
_spec_binned_400, dk, qz, off_spec = load_data('/SNS/users/m2d/git/reflectivity_ui/test/notebooks/data/REF_M_30889+30890_OffSpecBinned_Off_Off_v2_TOF_400.dat',
                                      index=42, q_axis=True, width=5)

plot_heatmap(dk, qz, np.log(off_spec), x_title='', y_title='', surface=False,
                 x_log=False, y_log=False)









    



(23688, 4)


File: /SNS/users/m2d/git/reflectivity_ui/test/notebooks/data/REF_M_30889+30890_OffSpecBinned_Off_Off_v2_TOF_400.dat
Binning: 84 282 [23688]
Scale = 10.7166
Const value = 0.000178571
Width = 5

Smooth dTOF=40 usec



In [8]:

    
_spec_40, dk, qz, off_spec = load_data('/SNS/users/m2d/git/reflectivity_ui/test/notebooks/data/REF_M_30889+30890_OffSpecSmooth_Off_Off_v2_TOF_40.dat',
                                      index=42, q_axis=True, width=5)

plot_heatmap(dk, qz, np.log(off_spec), x_title='', y_title='', surface=False,
                 x_log=False, y_log=False)









    



(23688, 3)


File: /SNS/users/m2d/git/reflectivity_ui/test/notebooks/data/REF_M_30889+30890_OffSpecSmooth_Off_Off_v2_TOF_40.dat
Binning: 84 282 [23688]
Scale = 8.62366
Const value = 0.000180723
Width = 5

Binned dTOF=40 usec



In [9]:

    
_spec_binned_40, dk, qz, off_spec = load_data('/SNS/users/m2d/git/reflectivity_ui/test/notebooks/data/REF_M_30889+30890_OffSpecBinned_Off_Off_v2_TOF_40.dat',
                                      index=42, q_axis=True, width=10)

plot_heatmap(dk, qz, np.log(off_spec), x_title='', y_title='', surface=False,
                 x_log=False, y_log=False)









    



(23688, 4)


File: /SNS/users/m2d/git/reflectivity_ui/test/notebooks/data/REF_M_30889+30890_OffSpecBinned_Off_Off_v2_TOF_40.dat
Binning: 84 282 [23688]
Scale = 10.6088
Const value = 0.000178571
Width = 10

Various ways of obtaining the specular reflectivity.

Curves labeled "Spec 2D" were obtained by summing the specular ridge in the off-specular map.
Curves labeled "Spec 1D" were obtained directly.
The number in brackets is the TOF



In [17]:

    
refl_400_cstq = np.loadtxt('/SNS/users/m2d/git/reflectivity_ui/test/notebooks/data/REF_M_30889+30890_Specular_Off_Off_CSTQ_v2_TOF_400.dat')
refl_40_cstq = np.loadtxt('/SNS/users/m2d/git/reflectivity_ui/test/notebooks/data/REF_M_30889+30890_Specular_Off_Off_CSTQ_v2_TOF_40.dat')
refl_400 = np.loadtxt('/SNS/users/m2d/git/reflectivity_ui/test/notebooks/data/REF_M_30889+30890_Specular_Off_Off_v2_TOF_400.dat')
refl_40 = np.loadtxt('/SNS/users/m2d/git/reflectivity_ui/test/notebooks/data/REF_M_30889+30890_Specular_Off_Off_v2_TOF_40.dat')

_scale = 1.0/2.8
_plot_list = [[_spec_400[0], _scale*_spec_400[1]],
              [_spec_40[0], _scale*_spec_40[1]],
              [_spec_binned_400[0], _scale*_spec_binned_400[1]],
              [_spec_binned_40[0], _scale*_spec_binned_40[1]],
             [refl_400_cstq.T[0], refl_400_cstq.T[1], refl_400_cstq.T[2]],
             [refl_40_cstq.T[0], refl_40_cstq.T[1], refl_40_cstq.T[2]],
             [refl_400.T[0], refl_400.T[1], refl_400.T[2]],
             [refl_40.T[0], refl_40.T[1], refl_40.T[2]],]
_plot_names = ["Spec 2D [400]", "Spec 2D [40]", "Spec 2D binned [400]", "Spec 2D binned [40]",
              "Spec 1D cst-q [400]", "Spec 1D cst-q [40]", "Spec 1D [400]", "Spec 1D [40]"]
plot1d(_plot_list, _plot_names,
           x_title='', y_title='',
           x_log=False, y_log=True, show_dx=True)