Merge pull request hiclib#14 from NelleV/select_partial_matrix

ENH added utility function to select subsample of the matrix

README.rst

@@ -1,15 +1,10 @@
 .. -*- mode: rst -*-
 
-|Travis|_ |AppVeyor|_ |Coveralls|_
+|Travis|_ |Coveralls|_
 
 .. |Travis| image:: https://api.travis-ci.org/hiclib/iced.png?branch=master
 .. _Travis: https://travis-ci.org/hiclib/iced
 
-.. |AppVeyor| image::
-   https://ci.appveyor.com/api/projects/status/github/hiclib/iced?branch=master&svg=true
-.. _AppVeyor:
-   https://ci.appveyor.com/project/sklearn-ci/scikit-learn/history
-
 .. |Coveralls| image::
    https://coveralls.io/repos/hiclib/iced/badge.svg?branch=master&service=github
 .. _Coveralls: https://coveralls.io/r/hiclib/iced

doc/modules/classes.rst

@@ -49,15 +49,13 @@ Functions
 
 .. _filter_ref:
 
-
 :mod:`iced.datasets`: Datasets
 ===============================================
 
 .. automodule:: iced.datasets
     :no-members:
     :no-inherited-members:
 
-
 Functions
 ---------
 .. currentmodule:: iced
@@ -72,3 +70,25 @@ Functions
 .. _datasets_ref:
 
 
+:mod:`iced.utils`: Utils
+===============================================
+
+.. automodule:: iced.utils
+    :no-members:
+    :no-inherited-members:
+
+
+Functions
+---------
+.. currentmodule:: iced
+
+.. autosummary::
+    :toctree: generated/
+    :template: function.rst
+
+
+    utils.get_intra_mask
+    utils.get_inter_mask
+    utils.extract_sub_contact_map
+
+.. _utils_ref:

doc/modules/utils.rst

@@ -0,0 +1,9 @@
+.. _utils:
+
+==============
+Utilities
+==============
+
+The :mod:`utils` submodule contains utilities function.
+
+.. currentmodule:: iced.utils

examples/utils/plot_extract_sample_map.py

@@ -0,0 +1,26 @@
+import matplotlib.pyplot as plt
+from matplotlib import colors
+
+from iced import datasets
+from iced.utils import extract_sub_contact_map
+
+"""
+Extracting parts of a contact map.
+
+This example shows how to extract contact counts associated to some chromosomes
+of the contact maps. Here, we extract chromosomes 1, 4 and 5 of the budding
+yeasts contact map
+"""
+
+# Loading a sample dataset
+counts, lengths = datasets.load_sample_yeast()
+sub_counts, sub_lengths = extract_sub_contact_map(counts, lengths, [0, 3, 4])
+
+fig, ax = plt.subplots()
+m = ax.matshow(sub_counts, cmap="Blues", norm=colors.SymLogNorm(1),
+               origin="bottom",
+               extent=(0, len(sub_counts), 0, len(sub_counts)))
+[ax.axhline(i, linewidth=1, color="#000000") for i in lengths.cumsum()]
+[ax.axvline(i, linewidth=1, color="#000000") for i in lengths.cumsum()]
+cb = fig.colorbar(m)
+ax.set_title("Chromosomes I, IV and V of yeast")

iced/datasets/base.py

@@ -46,7 +46,8 @@ def load_sample_yeast():
     Load and return a sample of S. cerevisiae contact count matrix from duan
     et al, Nature, 2009
 
-    Returns:
+    Returns
+    -------
         counts, lengths:
             tuple of two elements, the first a contact count matrix, the
             second an ndarray containing the lengths of the chromosomes.

iced/normalization.py

@@ -6,7 +6,7 @@
 
 def ICE_normalization(X, SS=None, max_iter=3000, eps=1e-4, copy=True,
                       norm='l1', verbose=0, output_bias=False,
-                      total_counts=None):
+                      total_counts=None, counts_profile=None):
     """
     ICE normalization
 
@@ -89,6 +89,8 @@ def ICE_normalization(X, SS=None, max_iter=3000, eps=1e-4, copy=True,
             raise NotImplementedError
 
         dbias = sum_ds.reshape((m, 1))
+        if counts_profile is not None:
+            dbias /= counts_profile[:, np.newaxis]
         # To avoid numerical instabilities
         dbias /= dbias[dbias != 0].mean()
 

iced/utils/_genome.py

@@ -1,4 +1,5 @@
 import numpy as np
+from .validation import is_symetric_or_tri
 
 
 def get_intra_mask(lengths):
@@ -71,6 +72,86 @@ def get_genomic_distances(lengths):
     return dis.astype(int)
 
 
+def extract_sub_contact_map(counts, lengths, chromosomes):
+    """
+    Extract the contact map associated to a given list of chromosome
+
+    Parameters
+    ----------
+    counts : ndarray (n, n)
+
+    lengths : ndarray (L, )
+
+    chromosomes : list of ids
+
+    Returns
+    -------
+    sub_counts, sub_lengths : (ndarray, ndarray)
+
+    Examples
+    --------
+
+    >>> from iced import datasets
+    >>> from iced.utils import extract_sub_contact_map
+    >>> counts, lengths = datasets.load_sample_yeast()
+    >>> scounts, slengths = extract_sub_contact_map(counts, lengths, [0, 2])
+    >>> print len(counts), len(scounts)
+    ... # doctest: +ELLIPSIS, +NORMALIZE_WHITESPACE
+    350 56
+    """
+    chromosomes = np.array(chromosomes)
+    if chromosomes.max() >= len(lengths):
+        raise ValueError(
+            "The chromosomes provided are not compatible with the "
+            "lengths array. Possible values are"
+            " %s" % " ".join("%s" % i for i in np.arange(len(lengths))))
+    if lengths.sum() != counts.shape[0]:
+        raise ValueError(
+            "The lengths provided is incompatible with the counts matrix"
+            "shape. The total lengths is %d while the contact count matrix "
+            "is %d" % (lengths.sum(), counts.shape[0]))
+
+    is_symetric_or_tri(counts)
+    chromosomes.sort()
+
+    new_lengths = lengths[chromosomes]
+    new_counts = np.zeros((new_lengths.sum(), new_lengths.sum()))
+    begin1, end1 = 0, 0
+    for i, l1 in enumerate(lengths):
+        end1 += l1
+        if i not in chromosomes:
+            begin1 = end1
+            continue
+        # Find position of this pair of chromosome in the matrix
+        new_num_chrom = (chromosomes == i).argmax()
+        if new_num_chrom == 0:
+            new_begin1 = 0
+        else:
+            new_begin1 = new_lengths.cumsum()[new_num_chrom - 1]
+        new_end1 = new_lengths.cumsum()[new_num_chrom]
+
+        begin2, end2 = 0, 0
+        for j, l2 in enumerate(lengths):
+            end2 += l2
+            if j not in chromosomes:
+                begin2 = end2
+                continue
+            # Find position of this pair of chromosome in the matrix
+            new_num_chrom = (chromosomes == j).argmax()
+            if new_num_chrom == 0:
+                new_begin2 = 0
+            else:
+                new_begin2 = new_lengths.cumsum()[new_num_chrom - 1]
+            new_end2 = new_lengths.cumsum()[new_num_chrom]
+            new_counts[new_begin1:new_end1,
+                       new_begin2:new_end2] = counts[begin1:end1, begin2:end2]
+            begin2 = end2
+
+        begin1 = end1
+
+    return new_counts, new_lengths
+
+
 def undersample_per_chr(X, lengths):
     """
     Undersample matrix to chromosomes

iced/utils/_validation.py

@@ -0,0 +1 @@
+

iced/utils/tests/test_genome.py

@@ -5,6 +5,7 @@
 from iced.utils._genome import get_inter_mask
 from iced.utils._genome import _change_lengths_resolution
 from iced.utils._genome import undersample_per_chr
+from iced.utils._genome import extract_sub_contact_map
 
 
 def test_get_intra_mask():
@@ -28,7 +29,7 @@ def test_get_inter_mask():
     true_mask = np.zeros((10, 10))
     true_mask[:5, :5] = 1
     true_mask[5:, 5:] = 1
-    assert_array_equal(mask, true_mask.astype(bool) == False)
+    assert_array_equal(mask, np.invert(true_mask.astype(bool)))
 
 
 def test_undersample_per_chr():
@@ -41,3 +42,13 @@ def test_undersample_per_chr():
     undersampled_X_true = np.array([[1, 0],
                                     [0, 0.5]])
     assert_array_equal(undersampled_X_true, undersampled_X)
+
+
+def test_return_sample():
+    lengths = np.array([50, 75])
+    n = lengths.sum()
+    X = np.random.randint(0, 50, (n, n))
+    X = np.triu(X)
+    sub_X, _ = extract_sub_contact_map(X, lengths, [0])
+    assert_array_equal(X[:lengths[0], :lengths[0]],
+                       sub_X)