Merge pull request #575 from GavinHuttley/develop

Miscellany

README.md

@@ -28,13 +28,25 @@ Our `cogent3.app` module provides a very different approach to using the library
 $ pip install cogent3
 ```
 
-### Installing with some other extra libraries, e.g. [plotly](https://github.com/plotly/plotly.py) and [pandas](https://github.com/pandas-dev/pandas) (optional)
+### Install `extra` -- adds visualisation support
+
+**NOTE:** Only available in the development version until release ≥2020.3
+
+The `extra` group includes python libraries required for visualisation (i.e. [plotly](https://pypi.org/project/plotly/) and [psutil](https://pypi.org/project/psutil/) plus [pandas](https://pypi.org/project/pandas/) (optional)
 
 ```bash
 $ pip install cogent3[extra]
 ```
 
-### Installing the development version (optional)
+### Install `dev` -- adds `cogent3` development related libraries
+
+The `dev` group includes python libraries required for development of `cogent3`.
+
+```bash
+$ pip install cogent3[dev]
+```
+
+### Install the development version
 
 ```bash
 $ pip install git+https://github.com/cogent3/cogent3.git@develop#egg=cogent3

setup.py

@@ -172,30 +172,30 @@ def CythonExtension(module_name, **kw):
     install_requires=["numpy", "scitrack", "tqdm", "tinydb"],
     extras_require={
         "dev": [
-            "pytest-azurepipelines",
-            "jupyterlab",
+            "black",
+            "click",
             "ipykernel",
             "ipywidgets",
-            "click",
-            "sphinx",
-            "sphinx_rtd_theme",
-            "sphinx-autobuild",
-            "sphinxcontrib-bibtex",
-            "numpydoc",
-            "nbformat",
-            "nbconvert",
+            "isort",
             "jupyter_client",
-            "ipykernel",
-            "nbsphinx",
+            "jupyterlab",
             "jupytext",
+            "nbconvert",
+            "nbformat",
+            "nbsphinx",
+            "numpydoc",
             "pandas",
             "plotly",
+            "psutil",
             "pytest",
+            "pytest-azurepipelines",
             "pytest-cov",
             "pytest>=4.3.0",
+            "sphinx",
+            "sphinx-autobuild",
+            "sphinx_rtd_theme",
+            "sphinxcontrib-bibtex",
             "tox",
-            "black",
-            "isort",
         ],
         "extra": ["pandas", "plotly", "psutil"],
     },

src/cogent3/maths/stats/number.py

@@ -1,4 +1,4 @@
-from collections import defaultdict
+from collections import Counter, defaultdict
 from collections.abc import Mapping, MutableMapping
 
 import numpy
@@ -204,6 +204,34 @@ def to_freqs(self):
         result = CategoryFreqs(self, total=self.sum)
         return result
 
+    def count(self, indices):
+        """
+        Parameters
+        ----------
+        indices
+            select element(s) from a multi-element tuple keys, must be int or
+            series of ints
+
+        Returns
+        -------
+        CategoryCounter
+        """
+        if isinstance(indices, int):
+            indices = [indices]
+
+        counts = Counter()
+        for key in self:
+            try:
+                sub_key = tuple(key[i] for i in indices)
+                sub_key = sub_key[0] if len(sub_key) == 1 else sub_key
+            except IndexError:
+                msg = f"indices {indices} too big for key {key}"
+                raise IndexError(msg)
+            counts[sub_key] += self[key]
+
+        result = self.__class__(data=counts)
+        return result
+
 
 class CategoryFreqs(MutableMapping, SummaryStatBase):
     """category frequencies with summary statistic attributes"""
@@ -287,6 +315,8 @@ def to_normalized(self):
 
 
 class NumberCounter(CategoryCounter):
+    """counts occurrences of numbers"""
+
     def __init__(self, data=None):
         super(NumberCounter, self).__init__(data=data)
 

tests/test_maths/test_stats/test_number.py

@@ -246,6 +246,63 @@ def test_numbers_update(self):
             counts = number.CategoryCounter("AAAACCCGGGGT")
             nums.update_from_counts(counts)
 
+    def test_count(self):
+        """correctly counts across key dimensions"""
+        data = [
+            ("T", "C"),
+            ("T", "T"),
+            ("T", "A"),
+            ("G", "A"),
+            ("G", "A"),
+            ("A", "C"),
+            ("A", "G"),
+            ("T", "T"),
+            ("T", "A"),
+            ("T", "T"),
+            ("A", "T"),
+            ("A", "C"),
+            ("A", "C"),
+            ("T", "A"),
+            ("A", "A"),
+            ("A", "C"),
+        ]
+        nums = number.CategoryCounter(data)
+        i0 = nums.count(0)
+        self.assertEqual(i0["T"], 7)
+        self.assertEqual(i0["G"], 2)
+        self.assertEqual(i0["A"], 7)
+        self.assertEqual(i0["C"], 0)
+        # works same if keys are strings
+        nums = number.CategoryCounter(["".join(e) for e in data])
+        i0 = nums.count(0)
+        self.assertEqual(i0["T"], 7)
+        self.assertEqual(i0["G"], 2)
+        self.assertEqual(i0["A"], 7)
+        self.assertEqual(i0["C"], 0)
+        with self.assertRaises(IndexError):
+            _ = nums.count([0, 3])
+
+        i0 = nums.count([0])
+        self.assertEqual(i0["G"], 2)
+        with self.assertRaises(IndexError):
+            _ = nums.count([0, 3])
+        i1 = nums.count(1)
+        self.assertEqual(i1["A"], 6)
+        self.assertEqual(i1["C"], 5)
+        self.assertEqual(i1["T"], 4)
+
+        data = {
+            ("A", "C", "G"): 10,
+            ("A", "T", "G"): 4,
+            ("C", "C", "G"): 3,
+            ("G", "C", "G"): 6,
+        }
+        nums = number.CategoryCounter(data)
+        i02 = nums.count([0, 2])
+        self.assertEqual(i02[("A", "G")], 14)
+        self.assertEqual(i02[("C", "G")], 3)
+        self.assertEqual(i02[("G", "G")], 6)
+
 
 if __name__ == "__main__":
     main()