Merge pull request #57 from loucerac/develop

Develop

docs/source/index.rst

@@ -12,12 +12,11 @@ and signaling circuits.
 Installation
 =============
 
-You can install drexml from PyPI using pip:
+You can install drexml using:
 
 .. code::
 
-   pip install git+ssh://[email protected]:loucerac/drexml.git
-
+   pip install git+https://github.com/loucerac/drexml.git@master
 
 
 Getting started

drexml/datasets.py

@@ -23,29 +23,30 @@ def load_physiological_circuits():
     return circuit_names.index[circuit_names["is_physiological"]].tolist()
 
 
-def fetch_file(disease, key, env, version="latest"):
-    """
-    Retrieve data.
-
-    Parameters:
+def fetch_file(key, env, version="latest"):
+    """Retrieve file from the environment.
 
-    - disease (str): The name of the disease.
-    - key (str): The key associated with the data.
-    - env (Union[str, pathlib.Path]): The environment variable or path containing the data.
-    - version (str, optional): The version of the data to retrieve (default: "latest").
-    - debug (bool, optional): Whether to enable debug mode (default: False).
-
-    Returns:
-    - frame (np.ndarray): The preprocessed data frame.
+    Parameters
+    ----------
+    key : str
+        Key of the file to retrieve.
+    env : dict
+        Environment.
+    version : str
+        Version of the file to retrieve.
 
-    Raises:
-    - ConnectTimeout: If a connection timeout occurs during retrieval.
+    Returns
+    -------
+    pathlib.Path
+        Path to the file.
 
+    Raises
+    ------
+    NotImplementedError
+        Not implemented yet.
     """
 
     print(f"Retrieving {key}")
-    experiment_env_path = pathlib.Path(disease)
-    env = read_disease_config(experiment_env_path)
     if env[key + "_zenodo"]:  # pragma: no cover
         if version == "latest":
             try:
@@ -352,10 +353,10 @@ def get_disease_data(disease):
     experiment_env_path = pathlib.Path(disease)
     env = read_disease_config(experiment_env_path)
 
-    gene_exp = fetch_file(disease, key="gene_exp", env=env, version="latest")
-    pathvals = fetch_file(disease, key="pathvals", env=env, version="latest")
-    circuits = fetch_file(disease, key="circuits", env=env, version="latest")
-    genes = fetch_file(disease, key="genes", env=env, version="latest")
+    gene_exp = fetch_file(key="gene_exp", env=env, version="latest")
+    pathvals = fetch_file(key="pathvals", env=env, version="latest")
+    circuits = fetch_file(key="circuits", env=env, version="latest")
+    genes = fetch_file(key="genes", env=env, version="latest")
 
     # gene_exp = gene_exp[genes.index[genes[genes_column]]]
 

drexml/explain.py

@@ -12,35 +12,39 @@
 from drexml.pystab import nogueria_test
 
 
-def matcorr(O, P):
+def matcorr(features, targets):
     """Fast correlation matrix computation.
 
     Parameters
     ----------
-    O : ndarray
+    features : ndarray [n_samples, n_features]
         A matrix of observations.
-    P : ndarray
+    targets : ndarray [n_samples, n_tasks]
         A matrix of predictions.
 
     Returns
     -------
     ndarray
-        The cross-correlation matrix.
+        The correlation matrix.
     """
-    n = O.shape[0]
+    n = features.shape[0]
 
-    DO = O - (
-        np.einsum("nt->t", O, optimize="optimal") / np.double(n)
-    )  # compute O - mean(O)
-    DP = P - (
-        np.einsum("nm->m", P, optimize="optimal") / np.double(n)
-    )  # compute P - mean(P)
+    features_center = features - (
+        np.einsum("nt->t", features, optimize="optimal") / np.double(n)
+    )
+    targets_center = targets - (
+        np.einsum("nm->m", targets, optimize="optimal") / np.double(n)
+    )
 
-    cov = np.einsum("nm,nt->mt", DP, DO, optimize="optimal")
+    cov = np.einsum("nm,nt->mt", targets_center, features_center, optimize="optimal")
 
-    varP = np.einsum("nm,nm->m", DP, DP, optimize="optimal")
-    varO = np.einsum("nt,nt->t", DO, DO, optimize="optimal")
-    tmp = np.einsum("m,t->mt", varP, varO, optimize="optimal")
+    targets_var = np.einsum(
+        "nm,nm->m", targets_center, targets_center, optimize="optimal"
+    )
+    features_var = np.einsum(
+        "nt,nt->t", features_center, features_center, optimize="optimal"
+    )
+    tmp = np.einsum("m,t->mt", targets_var, features_var, optimize="optimal")
 
     return cov / np.sqrt(tmp)
 

drexml/utils.py

@@ -50,7 +50,7 @@ def check_cli_arg_is_bool(arg):
     elif arg in ["false", "False", "FALSE", "0"]:
         arg = False
     else:
-        raise ValueError("debug must be a boolean")
+        raise ValueError(f"argument {arg} must be a boolean")
 
     return arg
 

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "drexml"
-version = "0.9.11"
+version = "0.9.12"
 description = "(DRExM³L) Drug REpurposing using and eXplainable Machine Learning and Mechanistic Models of signal transduction\""
 authors = [
     "Carlos Loucera <[email protected]>",