@@ -1,6 +1,7 @@ from numpy import * import unittest import scipy.optimize as opt +import copy as cp class SasData(object): """A data object for holding 1-d Q versus I SAS data. @@ -85,7 +86,69 @@ class SasData(object): for j in range(len(self)): out.i[j] = self.i[j] * other return out -#definition of the Guinier model for a sphere with a flat background + + ######### + # + # SasTrim Routines for Trimming and Masking SasData + # + ######### + + def make_mask(data_to_mask, mask_ranges): + """Generates a mask list of 0's and 1's to remove data in mask_ranges + + Takes a SasData object and creates a list with a 1 to 1 mapping with + SasData.q containing only zeros and ones based on the q-values in + mask_ranges. Mask_ranges takes the form of a list of length N of + lists of length two containing data to exclude. May expand in the future + to allow both positive and negative mask generation. + """ + + # need to make an assertion that mask ranges is a N x 2 set of lists + # not immediately sure how to do this at the moment + + # make a copy of the q list because we are going to change it + mask = cp.deepcopy(data_to_mask.q) + + for j in range(len(mask)): + + #for each pair of values in mask_ranges + for low, high in mask_ranges + + # check whether q values are within mask_ranges + if mask[j] >= low and q_mask[j] <= high: + mask[j] = 0 + break + else: + mask[j] = 1 + + self.mask_list = mask + + def mask(self): + """Applies a pre-calculated mask to a SasData object. + + Takes the pre-calculated mask from make_mask and creates a new + SasData object with the masked data points (those corresponding + to zeros in the mask) removed. The new object is then placed in + self.masked. In principle this should allow nested masking operations. + Whether this is a good idea remains to be seen. + """ + + assert type(self.mask) is list + assert len(self.mask) != 0 + + q_masked = extract(self.mask, self.q) + i_masked = extract(self.mask, self.i) + + self.masked = SasData(q_masked, i_masked) + return self.masked + + +################################################### +# +# Definition of data fitting models +# +################################################### + def guinier(q,param): """Function that returns a Guinier model @@ -121,6 +184,12 @@ def fit_guinier(data): return least_squares_fit +######################################## +# +# Unit tests +# +######################################## + class TestSasData(unittest.TestCase): def setUp(self): sas.py ff96dbf62a635e6e7d06319afbb77d2bc0f4e4dc Cameron Neylon cameron.neylon@stfc.ac.uk http://github.com/cameronneylon/sas/commit/1cf7f651cbcede5b536cc04341dbdcc6b5f88109 1cf7f651cbcede5b536cc04341dbdcc6b5f88109 2009-03-03T23:57:14-08:00 2009-03-03T23:57:14-08:00 added sastrim routines as methods of SasData- still need to write tests for these 2ba6a3f487b200f139f1351d52d4a32a335e3f77 Cameron Neylon cameron.neylon@stfc.ac.uk