Further renaming of variables

python/tests/beagle_numba.py

@@ -304,9 +304,9 @@ def compute_state_prob_matrix(fm, bm):
 def get_weights(typed_pos, untyped_pos, typed_cm, untyped_cm):
     """
     Get weights for the ungenotyped positions in a query haplotype, which are used in
-    linear interpolation of HMM state probabilities at the ungenotyped positions.
+    linear interpolation of hidden state probabilities at the ungenotyped positions.
 
-    In BB2016 (see below Equation 1), a weight between genotyped positions m and m + 1
+    In BB2016 (see below Equation 1), a weight between genotyped positions m and (m + 1)
     bounding ungenotyped marker x is denoted lambda_m,x.
 
     lambda_m,x = [g(m + 1) - g(x)] / [g(m + 1) - g(m)],
@@ -356,7 +356,7 @@ def interpolate_allele_probs(sm, ref_h, typed_pos, untyped_pos, typed_cm, untype
     where a is the number of alleles.
 
     Note that this function takes:
-    1. HMM state probability matrix at genotyped positions of size (m, h).
+    1. Hidden state probability matrix at genotyped positions of size (m, h).
     2. Reference haplotypes subsetted to ungenotyped positions of size (x, h).
 
     :param numpy.ndarray sm: HMM state probability matrix at genotyped positions.
@@ -365,7 +365,7 @@ def interpolate_allele_probs(sm, ref_h, typed_pos, untyped_pos, typed_cm, untype
     :param numpy.ndarray untyped_pos: Physical positions of ungenotyped markers (bp).
     :param numpy.ndarray typed_cm: Genetic map positions at genotyped markers (cM).
     :param numpy.ndarray untyped_cm: Genetic map positions at ungenotyped markers (cM).
-    :return: Interpolated allele probabilities.
+    :return: Imputed allele probabilities.
     :rtype: numpy.ndarray
     """
     # TODO: Allow for biallelic site matrix. Work with `_tskit.lshmm` properly.
@@ -403,15 +403,15 @@ def interpolate_allele_probs(sm, ref_h, typed_pos, untyped_pos, typed_cm, untype
 
 def get_map_alleles(allele_probs):
     """
-    Compute maximum a posteriori alleles at the ungenotyped markers of a query haplotype,
-    based on posterior marginal allele probabilities.
+    Compute maximum a posteriori (MAP) alleles at the ungenotyped markers
+    of a query haplotype, based on posterior marginal allele probabilities.
 
     The imputed alleles is an array of size x.
 
     WARN: If the allele probabilities are equal, then allele 0 is arbitrarily chosen.
     TODO: Investigate how often this happens and the effect of this arbitrary choice.
 
-    :param numpy.ndarray allele_probs: Interpolated allele probabilities.
+    :param numpy.ndarray allele_probs: Imputed allele probabilities.
     :return: Imputed alleles and their associated probabilities.
     :rtype: tuple(numpy.ndarray, numpy.ndarray)
     """
@@ -436,8 +436,8 @@ def run_interpolation_beagle(
 
     The physical positions of all the markers are an array of size (m + x).
 
-    This produces imputed alleles and their interpolated probabilities
-    at the ungenotyped markers of the query haplotype.
+    This produces imputed alleles and their probabilities at the ungenotyped markers
+    of the query haplotype.
 
     The default values of `ne` and `error_rate` are taken from BEAGLE 4.1, not 5.4.
     In BEAGLE 5.4, the default value of `ne` is 1e5 and `error_rate` is data-dependent.
@@ -448,7 +448,7 @@ def run_interpolation_beagle(
     :param int ne: Effective population size.
     :param float error_rate: Allele error rate.
     :param GeneticMap genetic_map: Genetic map.
-    :return: Imputed alleles and their interpolated probabilities.
+    :return: Imputed alleles and their probabilities.
     :rtype: tuple(numpy.ndarray, numpy.ndarray)
     """
     warnings.warn("This function is experimental and not fully tested.", stacklevel=1)
@@ -482,15 +482,15 @@ def run_interpolation_beagle(
     )
     sm = compute_state_prob_matrix(fm, bm)
     # Perform linear interpolation.
-    int_allele_probs = interpolate_allele_probs(
+    imputed_allele_probs = interpolate_allele_probs(
         sm=sm,
         ref_h=ref_h_untyped,
         typed_pos=typed_pos,
         untyped_pos=untyped_pos,
         typed_cm=typed_cm,
         untyped_cm=untyped_cm,
     )
-    imputed_alleles, max_allele_probs = get_map_alleles(int_allele_probs)
+    imputed_alleles, max_allele_probs = get_map_alleles(imputed_allele_probs)
     return (imputed_alleles, max_allele_probs)
 
 
@@ -511,8 +511,8 @@ def run_tsimpute(
 
     The physical positions of all the markers are an array of size (m + x).
 
-    This produces imputed alleles and their interpolated probabilities
-    at the ungenotyped markers of the query haplotype.
+    This produces imputed alleles and their probabilities at the ungenotyped markers
+    of the query haplotype.
 
     TODO: Set default precision. What should it be?
     TODO: Allow for imputation on user-specified genomic interval.
@@ -526,7 +526,7 @@ def run_tsimpute(
     :param numpy.ndarray mp: Per-site mismatch probabilities.
     :param int precision: Precision when running Li & Stephens HMM.
     :param GeneticMap genetic_map: Genetic map.
-    :return: Imputed alleles and their interpolated probabilities.
+    :return: Imputed alleles and their probabilities.
     :rtype: tuple(numpy.ndarray, numpy.ndarray)
     """
     warnings.warn(
@@ -565,15 +565,15 @@ def run_tsimpute(
     ls_hmm.backward_matrix(query_h_typed.T, fm.normalisation_factor, bm)
     sm = compute_state_prob_matrix(fm.decode(), bm.decode())
     # Perform linear interpolation.
-    int_allele_probs = interpolate_allele_probs(
+    imputed_allele_probs = interpolate_allele_probs(
         sm=sm,
         ref_h=ref_h_untyped,
         typed_pos=typed_pos,
         untyped_pos=untyped_pos,
         typed_cm=typed_cm,
         untyped_cm=untyped_cm,
     )
-    imputed_alleles, max_allele_probs = get_map_alleles(int_allele_probs)
+    imputed_alleles, max_allele_probs = get_map_alleles(imputed_allele_probs)
     return (imputed_alleles, max_allele_probs)