Major change of code base (started)

aaanalysis/__init__.py

@@ -1,10 +1,11 @@
 from aaanalysis.data_loader import load_dataset, load_scales
 from aaanalysis.aaclust import AAclust
-from aaanalysis.cpp import CPP, SequenceFeature, SplitRange
+from aaanalysis.cpp import CPP, CPPPlot, SequenceFeature, SplitRange
 from aaanalysis.dpulearn import dPULearn
-from aaanalysis.utils_plot import plot_settings, plot_set_legend, plot_gcfs
+from aaanalysis.plotting import plot_settings, plot_set_legend, plot_gcfs, plot_get_cmap, plot_get_cdict
 
 __all__ = ["load_dataset", "load_scales",
            "AAclust",
-           "CPP", "SequenceFeature", "SplitRange", "dPULearn",
-           "plot_settings", "plot_set_legend", "plot_gcfs"]
+           "CPP", "CPPPlot", "SequenceFeature", "SplitRange",
+           "dPULearn",
+           "plot_settings", "plot_set_legend", "plot_gcfs", "plot_get_cmap", "plot_get_cdict"]

aaanalysis/_utils.py → aaanalysis/_utils/_utils_check.py

@@ -1,41 +1,11 @@
-#! /usr/bin/python3
 """
-Config with folder structure
+Utility check functions
 """
-import os
-import platform
 from sklearn.utils import check_array, check_consistent_length
+import pandas as pd
 
 
-# Helper Function
-def _folder_path(super_folder, folder_name):
-    """Modification of separator (OS depending)"""
-    path = os.path.join(super_folder, folder_name + SEP)
-    return path
-
-
-# Folder
-SEP = "\\" if platform.system() == "Windows" else "/"
-FOLDER_PROJECT = os.path.dirname(os.path.abspath(__file__))
-FOLDER_DATA = _folder_path(FOLDER_PROJECT, 'data')
-URL_DATA = "https://github.com/breimanntools/aaanalysis/tree/master/aaanalysis/data/"
-
-# Default data for protein analysis
-STR_SCALES = "scales"   # Min-max normalized scales (from AAontology)
-STR_SCALES_RAW = "scales_raw"   # Ras scales (from AAontology)
-STR_SCALES_PC = "scales_pc"     # AAclust pc-based scales (pc: principal component)
-STR_SCALE_CAT = "scale_classification"  # AAontology
-STR_TOP60 = "top60"    # AAclustTop60
-STR_TOP60_EVAL = "top60_eval"  # AAclustTop60 evaluation
-
-# Column names
-COL_SCALE_ID = "scale_id"
-COL_SEQ = "sequence"
-COL_CAT = "category"
-COL_SUBCAT = "subcategory"
-
-
-# General check functions
+# Type checking functions
 def check_non_negative_number(name=None, val=None, min_val=0, max_val=None, accept_none=False,
                               just_int=True):
     """Check if value of given name variable is non-negative integer"""
@@ -108,8 +78,7 @@ def check_tuple(name=None, val=None, n=None):
         raise ValueError(error)
 
 
-# Data check functions
-# TODO update
+# Array checking functions
 def check_feat_matrix(X=None, names=None, labels=None):
     """Check if X and y match (y can be labels or names). Otherwise, transpose X or give error."""
     # TODO type check
@@ -141,7 +110,40 @@ def check_feat_matrix(X=None, y=None):
 """
 
 
-# Plotting & print functions
-def print_red(input_str, **args):
-    """Prints the given string in red text."""
-    print(f"\033[91m{input_str}\033[0m", **args)
+# df checking functions
+def check_col_in_df(df=None, name_df=None, col=None, col_type=None, accept_nan=False, error_if_exists=False):
+    """
+    Check if the column exists in the DataFrame, if the values have the correct type, and if NaNs are allowed.
+    """
+    # Check if the column already exists and raise error if error_if_exists is True
+    if error_if_exists and (col in df.columns):
+        raise ValueError(f"Column '{col}' already exists in '{name_df}'")
+
+    # Check if the column exists in the DataFrame
+    if col not in df.columns:
+        raise ValueError(f"'{col}' must be a column in '{name_df}': {list(df.columns)}")
+
+    # Make col_type a list if it is not already
+    if col_type is not None and not isinstance(col_type, list):
+        col_type = [col_type]
+
+    # Check if the types match
+    if col_type is not None:
+        wrong_types = [x for x in df[col] if not any([isinstance(x, t) for t in col_type])]
+
+        # Remove NaNs from the list of wrong types if they are accepted
+        if accept_nan:
+            wrong_types = [x for x in wrong_types if not pd.isna(x)]
+
+        if len(wrong_types) > 0:
+            raise ValueError(f"Values in '{col}' should be of type(s) {col_type}, "
+                             f"but the following values do not match: {wrong_types}")
+
+    # Check if NaNs are present when they are not accepted
+    if not accept_nan:
+        if df[col].isna().sum() > 0:
+            raise ValueError(f"NaN values are not allowed in '{col}'.")
+
+
+
+

aaanalysis/_utils/_utils_constants.py

@@ -0,0 +1,71 @@
+"""
+This is a script for constants (e.g., names or column names)
+"""
+
+
+# Default datasets for protein analysis
+STR_SCALES = "scales"   # Min-max normalized scales (from AAontology)
+STR_SCALES_RAW = "scales_raw"   # Raw scales (from AAontology)
+STR_SCALES_PC = "scales_pc"     # AAclust pc-based scales (pc: principal component)
+STR_SCALE_CAT = "scale_classification"  # AAontology
+STR_TOP60 = "top60"    # AAclustTop60
+STR_TOP60_EVAL = "top60_eval"  # AAclustTop60 evaluation
+
+# Column name datasets (DOM_GSEC)
+COL_ENTRY = "entry"     # ACC, protein entry, uniprot id
+COL_NAME = "name"       # Entry name, Protein name, Uniprot Name
+COL_LABEL = "label"
+COL_SEQ = "sequence"
+COLS_PARTS = ["jmd_n", "tmd", "jmd_c"]
+COL_TMD_START = "tmd_start"
+COL_TMD_STOP = "tmd_stop"
+COLS_SEQ_INFO = [COL_SEQ, COL_TMD_START, COL_TMD_STOP]  # TODO
+
+# Column names scales (scale_classification)
+COL_SCALE_ID = "scale_id"
+COL_CAT = "category"
+COL_SUBCAT = "subcategory"
+COL_SCALE_NAME = "scale_name"
+COL_SCALE_DES = "scale_description"
+
+# Column names cpp features
+COL_FEATURE = "feature"
+# COL_CAT, COL_SUBCAT, COL_SCALE_NAME, COL_SCALE_DES
+COL_ABS_AUC = "abs_auc"
+COL_ABS_MEAN_DIF = "abs_mean_dif"
+COL_MEAN_DIF = "mean_dif"
+COL_STD_TEST = "std_test"
+COL_STD_REF = "std_ref"
+COL_PVAL_MW = "p_val_mann_whitney"
+COL_PVAL_FDR = "p_val_fdr_bh"
+COL_POSITION = "positions"
+
+COL_FEAT_IMPORTANCE = "feat_importance"
+COO_FEAT_IMP_STD = "feat_importance_std"
+COL_FEAT_IMPACT = "feat_impact"
+
+# Standard colors
+COLOR_SHAP_POS = '#FF0D57'  # (255, 13, 87)
+COLOR_SHAP_NEG = '#1E88E5'  # (30, 136, 229)
+COLOR_FEAT_POS = '#9D2B39'  # (157, 43, 57) Mean difference
+COLOR_FEAT_NEG = '#326599'  # (50, 101, 133) Mean difference
+COLOR_FEAT_IMP = '#7F7F7F'  # (127, 127, 127) feature importance
+COLOR_TMD = '#00FA9A'       # (0, 250, 154)
+COLOR_JMD = '#0000FF'       # (0, 0, 255)
+
+DICT_COLOR = {"SHAP_POS": COLOR_SHAP_POS,
+              "SHAP_NEG": COLOR_SHAP_NEG,
+              "FEAT_POS": COLOR_FEAT_POS,
+              "FEAT_NEG": COLOR_FEAT_NEG,
+              "FEAT_IMP": COLOR_FEAT_IMP,
+              "TMD": COLOR_TMD,
+              "JMD": COLOR_JMD}
+
+DICT_COLOR_CAT = {"ASA/Volume": "tab:blue",
+                  "Composition": "tab:orange",
+                  "Conformation": "tab:green",
+                  "Energy": "tab:red",
+                  "Others": "tab:gray",
+                  "Polarity": "gold",
+                  "Shape": "tab:cyan",
+                  "Structure-Activity": "tab:brown"}

aaanalysis/_utils/_utils_output.py

@@ -0,0 +1,32 @@
+"""
+This is a script for adjust output (mainly for python console)
+"""
+import numpy as np
+
+
+# I Helper Functions
+# Plotting & print functions
+def print_red(input_str, **args):
+    """Prints the given string in red text."""
+    print(f"\033[91m{input_str}\033[0m", **args)
+
+
+# Progress bar
+def print_start_progress():
+    """Print start progress"""
+    progress_bar = " " * 25
+    print_red(f"\r   |{progress_bar}| 0.00%", end="")
+
+
+def print_progress(i=0, n=0):
+    """Print progress"""
+    progress = min(np.round(i/n * 100, 2), 100)
+    progress_bar = "#" * int(progress/4) + " " * (25-int(progress/4))
+    print_red(f"\r   |{progress_bar}| {progress:.2f}%", end="")
+
+
+def print_finished_progress():
+    """Print finished progress bar"""
+    progress_bar = "#" * 25
+    print_red(f"\r   |{progress_bar}| 100.00%")
+

aaanalysis/aaclust/_utils.py → aaanalysis/_utils/utils_aaclust.py

@@ -1,6 +1,5 @@
-#! /usr/bin/python3
 """
-Config with folder structure
+Utility functions for AAclust module
 """
 import inspect