pycompound 0.1.2__tar.gz → 0.1.4__tar.gz

{pycompound-0.1.2/src/pycompound.egg-info → pycompound-0.1.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pycompound
-Version: 0.1.2
+Version: 0.1.4
 Summary: Python package to perform compound identification in mass spectrometry via spectral library matching.
 Author-email: Hunter Dlugas <fy7392@wayne.edu>
 License-Expression: MIT

{pycompound-0.1.2 → pycompound-0.1.4}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "pycompound"
-version = "0.1.2"
+version = "0.1.4"
 authors = [
   { name="Hunter Dlugas", email="fy7392@wayne.edu" },
 ]

{pycompound-0.1.2 → pycompound-0.1.4}/src/app.py

@@ -25,7 +25,6 @@ import ast
 from numbers import Real
 
 
-
 _LOG_QUEUE: asyncio.Queue[str] = asyncio.Queue()
 
 def _run_with_redirects(fn, writer, *args, **kwargs):
@@ -474,8 +473,15 @@ def run_parameter_tuning_ui(platform: str):
 
 
 
+'''
+app_ui = ui.page_fluid(
+    ui.output_ui("main_ui"),
+    ui.output_text("status_output")
+)
+'''
 
 app_ui = ui.page_fluid(
+    ui.head_content(ui.tags.link(rel="icon", href="emblem.png")),
     ui.output_ui("main_ui"),
     ui.output_text("status_output")
 )
@@ -688,7 +694,6 @@ def server(input, output, session):
         img: ImgData = {"src": str(dir / "www/emblem.png"), "width": "320px", "height": "250px"}
         return img
 
-
     @output
     @render.ui
     def main_ui():
@@ -697,6 +702,7 @@ def server(input, output, session):
                 ui.h2("Main Menu"),
                 ui.div(
                     ui.output_image("image"),
+                    #ui.img(src="emblem.png", width="320px", height="250px"),
                     style=(
                         "position:fixed; top:0; left:50%; transform:translateX(-50%); "
                         "z-index:1000; text-align:center; padding:10px; background-color:white;"
@@ -884,11 +890,18 @@ def server(input, output, session):
         weights = [float(weight.strip()) for weight in input.weights().split(",") if weight.strip()]
         weights = {'Cosine':weights[0], 'Shannon':weights[1], 'Renyi':weights[2], 'Tsallis':weights[3]}
 
+        high_quality_reference_library_tmp2 = False
+        if input.high_quality_reference_library() != 'False':
+            high_quality_reference_library_tmp2 = True
+
+        print(input.high_quality_reference_library())
+        print(high_quality_reference_library_tmp2)
+
         if input.chromatography_platform() == "HRMS":
-            fig = generate_plots_on_HRMS_data(query_data=input.query_data()[0]['datapath'], reference_data=input.reference_data()[0]['datapath'], spectrum_ID1=spectrum_ID1, spectrum_ID2=spectrum_ID2, similarity_measure=input.similarity_measure(), weights=weights, spectrum_preprocessing_order=input.spectrum_preprocessing_order(), high_quality_reference_library=input.high_quality_reference_library(), mz_min=input.mz_min(), mz_max=input.mz_max(), int_min=input.int_min(), int_max=input.int_max(), window_size_centroiding=input.window_size_centroiding(), window_size_matching=input.window_size_matching(), noise_threshold=input.noise_threshold(), wf_mz=input.wf_mz(), wf_intensity=input.wf_int(), LET_threshold=input.LET_threshold(), entropy_dimension=input.entropy_dimension(), y_axis_transformation=input.y_axis_transformation(), return_plot=True)
+            fig = generate_plots_on_HRMS_data(query_data=input.query_data()[0]['datapath'], reference_data=input.reference_data()[0]['datapath'], spectrum_ID1=spectrum_ID1, spectrum_ID2=spectrum_ID2, similarity_measure=input.similarity_measure(), weights=weights, spectrum_preprocessing_order=input.spectrum_preprocessing_order(), high_quality_reference_library=high_quality_reference_library_tmp2, mz_min=input.mz_min(), mz_max=input.mz_max(), int_min=input.int_min(), int_max=input.int_max(), window_size_centroiding=input.window_size_centroiding(), window_size_matching=input.window_size_matching(), noise_threshold=input.noise_threshold(), wf_mz=input.wf_mz(), wf_intensity=input.wf_int(), LET_threshold=input.LET_threshold(), entropy_dimension=input.entropy_dimension(), y_axis_transformation=input.y_axis_transformation(), return_plot=True)
             plt.show()
         elif input.chromatography_platform() == "NRMS":
-            fig = generate_plots_on_NRMS_data(query_data=input.query_data()[0]['datapath'], reference_data=input.reference_data()[0]['datapath'], spectrum_ID1=spectrum_ID1, spectrum_ID2=spectrum_ID2, similarity_measure=input.similarity_measure(), spectrum_preprocessing_order=input.spectrum_preprocessing_order(), high_quality_reference_library=input.high_quality_reference_library(), mz_min=input.mz_min(), mz_max=input.mz_max(), int_min=input.int_min(), int_max=input.int_max(), noise_threshold=input.noise_threshold(), wf_mz=input.wf_mz(), wf_intensity=input.wf_int(), LET_threshold=input.LET_threshold(), entropy_dimension=input.entropy_dimension(), y_axis_transformation=input.y_axis_transformation(), return_plot=True)
+            fig = generate_plots_on_NRMS_data(query_data=input.query_data()[0]['datapath'], reference_data=input.reference_data()[0]['datapath'], spectrum_ID1=spectrum_ID1, spectrum_ID2=spectrum_ID2, similarity_measure=input.similarity_measure(), spectrum_preprocessing_order=input.spectrum_preprocessing_order(), high_quality_reference_library=high_quality_reference_library_tmp2, mz_min=input.mz_min(), mz_max=input.mz_max(), int_min=input.int_min(), int_max=input.int_max(), noise_threshold=input.noise_threshold(), wf_mz=input.wf_mz(), wf_intensity=input.wf_int(), LET_threshold=input.LET_threshold(), entropy_dimension=input.entropy_dimension(), y_axis_transformation=input.y_axis_transformation(), return_plot=True)
             plt.show()
         with io.BytesIO() as buf:
             fig.savefig(buf, format="png", dpi=150, bbox_inches="tight")
@@ -1115,3 +1128,26 @@ def server(input, output, session):
 app = App(app_ui, server)
 
 
+
+'''
+from starlette.middleware.base import BaseHTTPMiddleware
+from starlette.requests import Request
+
+class _InjectBaseURLMiddleware(BaseHTTPMiddleware):
+    async def dispatch(self, request: Request, call_next):
+        hdrs = dict(request.scope.get("headers", []))
+        if b"rstudio-connect-app-base-url" not in hdrs:
+            host = request.headers.get("x-forwarded-host") or request.headers.get("host") or ""
+            proto = request.headers.get("x-forwarded-proto") or "https"
+            root_path = (request.scope.get("root_path") or "").rstrip("/")
+            base = f"{proto}://{host}{root_path}"
+            new_headers = list(request.scope.get("headers", [])) + [
+                (b"rstudio-connect-app-base-url", base.encode("utf-8"))
+            ]
+            request.scope["headers"] = new_headers
+        return await call_next(request)
+
+app.starlette_app.add_middleware(_InjectBaseURLMiddleware)
+'''
+
+

{pycompound-0.1.2 → pycompound-0.1.4}/src/pycompound/plot_spectra.py

@@ -92,8 +92,8 @@ def generate_plots_on_HRMS_data(query_data=None, reference_data=None, spectrum_I
         print(f'Error: spectrum_preprocessing_order must contain only \'C\', \'F\', \'M\', \'N\', \'L\', \'W\'.')
         sys.exit()
 
-    if similarity_measure not in ['cosine','shannon','renyi','tsallis','mixture','jaccard','dice','3w_jaccard','sokal_sneath','binary_cosine','mountford','mcconnaughey','driver_kroeber','simpson','braun_banquet','fager_mcgowan','kulczynski','interection','hamming','hellinger']:
-        print('\nError: similarity_measure must be either cosine, shannon, renyi, tsallis, mixture, jaccard, dice, 3w_jaccard, sokal_sneath, binary_cosine, mountford, mcconnaughey, driver_kroeber, simpson, braun_banquet, fager_mcgowan, kulczynski, interection, hamming, or hellinger.')
+    if similarity_measure not in ['cosine','shannon','renyi','tsallis','mixture','jaccard','dice','3w_jaccard','sokal_sneath','binary_cosine','mountford','mcconnaughey','driver_kroeber','simpson','braun_banquet','fager_mcgowan','kulczynski','intersection','hamming','hellinger']:
+        print('\nError: similarity_measure must be either cosine, shannon, renyi, tsallis, mixture, jaccard, dice, 3w_jaccard, sokal_sneath, binary_cosine, mountford, mcconnaughey, driver_kroeber, simpson, braun_banquet, fager_mcgowan, kulczynski, intersection, hamming, or hellinger.')
         sys.exit()
 
     if isinstance(int_min,int) is True:
@@ -243,10 +243,12 @@ def generate_plots_on_HRMS_data(query_data=None, reference_data=None, spectrum_I
             r_spec[:,1] = LE_transform(r_spec[:,1], LET_threshold, normalization_method=normalization_method)
         if transformation == 'N' and q_spec.shape[0] > 1 and r_spec.shape[1] > 1:
             q_spec = remove_noise(q_spec, nr = noise_threshold)
-            r_spec = remove_noise(r_spec, nr = noise_threshold)
+            if high_quality_reference_library == False or high_quality_reference_library == 'False':
+                r_spec = remove_noise(r_spec, nr = noise_threshold)
         if transformation == 'F' and q_spec.shape[0] > 1 and r_spec.shape[1] > 1:
             q_spec = filter_spec_lcms(q_spec, mz_min = mz_min, mz_max = mz_max, int_min = int_min, int_max = int_max, is_matched = is_matched)
-            r_spec = filter_spec_lcms(r_spec, mz_min = mz_min, mz_max = mz_max, int_min = int_min, int_max = int_max, is_matched = is_matched)
+            if high_quality_reference_library == False or high_quality_reference_library == 'False':
+                r_spec = filter_spec_lcms(r_spec, mz_min = mz_min, mz_max = mz_max, int_min = int_min, int_max = int_max, is_matched = is_matched)
 
     q_ints = q_spec[:,1]
     r_ints = r_spec[:,1]
@@ -291,12 +293,15 @@ def generate_plots_on_HRMS_data(query_data=None, reference_data=None, spectrum_I
         plt.yticks([])
 
 
+    print('\n\n\n')
+    print(high_quality_reference_library)
+    print('\n\n\n')
     plt.subplots_adjust(top=0.8, hspace=0.92, bottom=0.3)
     plt.figlegend(loc = 'upper center')
     fig.text(0.05, 0.18, f'Similarity Measure: {similarity_measure.capitalize()}', fontsize=7)
     fig.text(0.05, 0.15, f'Similarity Score: {round(similarity_score,4)}', fontsize=7)
     fig.text(0.05, 0.12, f"Spectrum Preprocessing Order: {''.join(spectrum_preprocessing_order)}", fontsize=7)
-    fig.text(0.05, 0.09, f'High Quality Reference Library: {high_quality_reference_library}', fontsize=7)
+    fig.text(0.05, 0.09, f'High Quality Reference Library: {str(high_quality_reference_library)}', fontsize=7)
     fig.text(0.05, 0.06, f'Window Size (Centroiding): {window_size_centroiding}', fontsize=7)
     fig.text(0.05, 0.03, f'Window Size (Matching): {window_size_matching}', fontsize=7)
     fig.text(0.45, 0.18, f'Raw-Scale M/Z Range: [{mz_min_tmp},{mz_max_tmp}]', fontsize=7)
@@ -387,8 +392,8 @@ def generate_plots_on_NRMS_data(query_data=None, reference_data=None, spectrum_I
         print(f'Error: spectrum_preprocessing_order must contain only \'F\', \'N\', \'W\', \'L\'.')
         sys.exit()
 
-    if similarity_measure not in ['cosine','shannon','renyi','tsallis','mixture','jaccard','dice','3w_jaccard','sokal_sneath','binary_cosine','mountford','mcconnaughey','driver_kroeber','simpson','braun_banquet','fager_mcgowan','kulczynski','interection','hamming','hellinger']:
-        print('\nError: similarity_measure must be either cosine, shannon, renyi, tsallis, mixture, jaccard, dice, 3w_jaccard, sokal_sneath, binary_cosine, mountford, mcconnaughey, driver_kroeber, simpson, braun_banquet, fager_mcgowan, kulczynski, interection, hamming, or hellinger.')
+    if similarity_measure not in ['cosine','shannon','renyi','tsallis','mixture','jaccard','dice','3w_jaccard','sokal_sneath','binary_cosine','mountford','mcconnaughey','driver_kroeber','simpson','braun_banquet','fager_mcgowan','kulczynski','intersection','hamming','hellinger']:
+        print('\nError: similarity_measure must be either cosine, shannon, renyi, tsallis, mixture, jaccard, dice, 3w_jaccard, sokal_sneath, binary_cosine, mountford, mcconnaughey, driver_kroeber, simpson, braun_banquet, fager_mcgowan, kulczynski, intersection, hamming, or hellinger.')
         sys.exit()
 
     if isinstance(int_min,int) is True:
@@ -528,11 +533,11 @@ def generate_plots_on_NRMS_data(query_data=None, reference_data=None, spectrum_I
             r_spec[:,1] = LE_transform(r_spec[:,1], LET_threshold, normalization_method)
         if transformation == 'N':
             q_spec = remove_noise(q_spec, nr = noise_threshold)
-            if high_quality_reference_library == False:
+            if high_quality_reference_library == False or high_quality_reference_library == 'False':
                 r_spec = remove_noise(r_spec, nr = noise_threshold)
         if transformation == 'F':
             q_spec = filter_spec_gcms(q_spec, mz_min = mz_min, mz_max = mz_max, int_min = int_min, int_max = int_max)
-            if high_quality_reference_library == False:
+            if high_quality_reference_library == False or high_quality_reference_library == 'False':
                 r_spec = filter_spec_gcms(r_spec, mz_min = mz_min, mz_max = mz_max, int_min = int_min, int_max = int_max)
 
     if q_spec.shape[0] > 1:
@@ -580,7 +585,7 @@ def generate_plots_on_NRMS_data(query_data=None, reference_data=None, spectrum_I
     fig.text(0.05, 0.15, f'Similarity Measure: {similarity_measure.capitalize()}', fontsize=7)
     fig.text(0.05, 0.12, f'Similarity Score: {round(similarity_score,4)}', fontsize=7)
     fig.text(0.05, 0.09, f"Spectrum Preprocessing Order: {''.join(spectrum_preprocessing_order)}", fontsize=7)
-    fig.text(0.05, 0.06, f'High Quality Reference Library: {high_quality_reference_library}', fontsize=7)
+    fig.text(0.05, 0.06, f'High Quality Reference Library: {str(high_quality_reference_library)}', fontsize=7)
     fig.text(0.05, 0.03, f'Raw-Scale M/Z Range: [{min_mz},{max_mz}]', fontsize=7)
     fig.text(0.45, 0.15, f'Raw-Scale Intensity Range: [{int_min_tmp},{int_max_tmp}]', fontsize=7)
     fig.text(0.45, 0.12, f'Noise Threshold: {noise_threshold}', fontsize=7)

{pycompound-0.1.2 → pycompound-0.1.4}/src/pycompound/spec_lib_matching.py

@@ -37,6 +37,7 @@ def _eval_one_HRMS(df_query, df_reference, unique_query_ids, unique_reference_id
         LET_threshold=LET_threshold_tmp,
         entropy_dimension=entropy_dimension_tmp,
         high_quality_reference_library=high_quality_reference_library_tmp,
+        verbose=True
     )
 
     return (
@@ -441,21 +442,25 @@ def tune_params_on_NRMS_data_shiny(query_data=None, reference_data=None, grid=No
 
 
 
-def get_acc_HRMS(df_query, df_reference, unique_query_ids, unique_reference_ids, similarity_measure, weights, spectrum_preprocessing_order, mz_min, mz_max, int_min, int_max, window_size_centroiding, window_size_matching, noise_threshold, wf_mz, wf_int, LET_threshold, entropy_dimension, high_quality_reference_library):
+def get_acc_HRMS(df_query, df_reference, unique_query_ids, unique_reference_ids, similarity_measure, weights, spectrum_preprocessing_order, mz_min, mz_max, int_min, int_max, window_size_centroiding, window_size_matching, noise_threshold, wf_mz, wf_int, LET_threshold, entropy_dimension, high_quality_reference_library, verbose=True):
 
     n_top_matches_to_save = 1
 
     all_similarity_scores =  []
     for query_idx in range(0,len(unique_query_ids)):
-        print(f'query spectrum #{query_idx} is being identified')
+        if verbose is True:
+            print(f'query spectrum #{query_idx} is being identified')
         q_idxs_tmp = np.where(df_query.iloc[:,0] == unique_query_ids[query_idx])[0]
         q_spec_tmp = np.asarray(pd.concat([df_query.iloc[q_idxs_tmp,1], df_query.iloc[q_idxs_tmp,2]], axis=1).reset_index(drop=True))
+        #q_spec_tmp = q_spec_tmp.astype(float)
 
         similarity_scores = []
         for ref_idx in range(0,len(unique_reference_ids)):
             q_spec = q_spec_tmp
             r_idxs_tmp = np.where(df_reference.iloc[:,0] == unique_reference_ids[ref_idx])[0]
             r_spec = np.asarray(pd.concat([df_reference.iloc[r_idxs_tmp,1], df_reference.iloc[r_idxs_tmp,2]], axis=1).reset_index(drop=True))
+            #print(r_spec)
+            #r_spec = r_spec.astype(float)
 
             is_matched = False
             for transformation in spectrum_preprocessing_order:
@@ -529,7 +534,7 @@ def get_acc_HRMS(df_query, df_reference, unique_query_ids, unique_reference_ids,
 
 
 
-def get_acc_NRMS(df_query, df_reference, unique_query_ids, unique_reference_ids, similarity_measure, weights, spectrum_preprocessing_order, mz_min, mz_max, int_min, int_max, noise_threshold, wf_mz, wf_int, LET_threshold, entropy_dimension, high_quality_reference_library):
+def get_acc_NRMS(df_query, df_reference, unique_query_ids, unique_reference_ids, similarity_measure, weights, spectrum_preprocessing_order, mz_min, mz_max, int_min, int_max, noise_threshold, wf_mz, wf_int, LET_threshold, entropy_dimension, high_quality_reference_library, verbose=True):
 
     n_top_matches_to_save = 1
 
@@ -546,7 +551,7 @@ def get_acc_NRMS(df_query, df_reference, unique_query_ids, unique_reference_ids,
         similarity_scores = []
         for ref_idx in range(0,len(unique_reference_ids)):
             q_spec = q_spec_tmp
-            if ref_idx % 1000 == 0:
+            if verbose is True and ref_idx % 1000 == 0:
                 print(f'Query spectrum #{query_idx} has had its similarity with {ref_idx} reference library spectra computed')
             r_idxs_tmp = np.where(df_reference.iloc[:,0] == unique_reference_ids[ref_idx])[0]
             r_spec_tmp = np.asarray(pd.concat([df_reference.iloc[r_idxs_tmp,1], df_reference.iloc[r_idxs_tmp,2]], axis=1).reset_index(drop=True))
@@ -615,7 +620,7 @@ def get_acc_NRMS(df_query, df_reference, unique_query_ids, unique_reference_ids,
 
 
 
-def run_spec_lib_matching_on_HRMS_data(query_data=None, reference_data=None, likely_reference_ids=None, similarity_measure='cosine', weights={'Cosine':0.25,'Shannon':0.25,'Renyi':0.25,'Tsallis':0.25}, spectrum_preprocessing_order='FCNMWL', high_quality_reference_library=False, mz_min=0, mz_max=9999999, int_min=0, int_max=9999999, window_size_centroiding=0.5, window_size_matching=0.5, noise_threshold=0.0, wf_mz=0.0, wf_intensity=1.0, LET_threshold=0.0, entropy_dimension=1.1, n_top_matches_to_save=1, print_id_results=False, output_identification=None, output_similarity_scores=None, return_ID_output=False):
+def run_spec_lib_matching_on_HRMS_data(query_data=None, reference_data=None, likely_reference_ids=None, similarity_measure='cosine', weights={'Cosine':0.25,'Shannon':0.25,'Renyi':0.25,'Tsallis':0.25}, spectrum_preprocessing_order='FCNMWL', high_quality_reference_library=False, mz_min=0, mz_max=9999999, int_min=0, int_max=9999999, window_size_centroiding=0.5, window_size_matching=0.5, noise_threshold=0.0, wf_mz=0.0, wf_intensity=1.0, LET_threshold=0.0, entropy_dimension=1.1, n_top_matches_to_save=1, print_id_results=False, output_identification=None, output_similarity_scores=None, return_ID_output=False, verbose=True):
     '''
     runs spectral library matching on high-resolution mass spectrometry (HRMS) data
 
@@ -762,14 +767,13 @@ def run_spec_lib_matching_on_HRMS_data(query_data=None, reference_data=None, lik
 
     all_similarity_scores =  []
     for query_idx in range(0,len(unique_query_ids)):
-        print(f'query spectrum #{query_idx} is being identified')
+        if verbose is True:
+            print(f'query spectrum #{query_idx} is being identified')
         q_idxs_tmp = np.where(df_query.iloc[:,0] == unique_query_ids[query_idx])[0]
         q_spec_tmp = np.asarray(pd.concat([df_query.iloc[q_idxs_tmp,1], df_query.iloc[q_idxs_tmp,2]], axis=1).reset_index(drop=True))
 
         similarity_scores = []
         for ref_idx in range(0,len(unique_reference_ids)):
-            #if ref_idx % 100 == 0:
-            #    print(f'Query spectrum #{query_idx} has had its similarity with {ref_idx} reference library spectra computed')
             q_spec = q_spec_tmp
             r_idxs_tmp = np.where(df_reference.iloc[:,0] == unique_reference_ids[ref_idx])[0]
             r_spec = np.asarray(pd.concat([df_reference.iloc[r_idxs_tmp,1], df_reference.iloc[r_idxs_tmp,2]], axis=1).reset_index(drop=True))
@@ -1008,9 +1012,9 @@ def run_spec_lib_matching_on_NRMS_data(query_data=None, reference_data=None, lik
 
         similarity_scores = []
         for ref_idx in range(0,len(unique_reference_ids)):
-            q_spec = q_spec_tmp
-            if ref_idx % 1000 == 0:
+            if verbose is True and ref_idx % 1000 == 0:
                 print(f'Query spectrum #{query_idx} has had its similarity with {ref_idx} reference library spectra computed')
+            q_spec = q_spec_tmp
             r_idxs_tmp = np.where(df_reference.iloc[:,0] == unique_reference_ids[ref_idx])[0]
             r_spec_tmp = np.asarray(pd.concat([df_reference.iloc[r_idxs_tmp,1], df_reference.iloc[r_idxs_tmp,2]], axis=1).reset_index(drop=True))
             r_spec = convert_spec(r_spec_tmp,mzs)

pycompound-0.1.4/src/pycompound/tuning_CLI_DE.py

@@ -0,0 +1,233 @@
+
+#!/usr/bin/env python3
+import argparse
+import sys
+import json
+from pathlib import Path
+from typing import Dict, List, Tuple
+import numpy as np
+import pandas as pd
+from scipy.optimize import differential_evolution
+from pycompound.spec_lib_matching import get_acc_HRMS, get_acc_NRMS
+
+
+ALL_PARAMS = [
+    "window_size_centroiding",
+    "window_size_matching",
+    "noise_threshold",
+    "wf_mz",
+    "wf_int",
+    "LET_threshold",
+    "entropy_dimension"
+]
+
+SUGGESTED_BOUNDS = {
+    "window_size_centroiding": (0.0, 0.5),
+    "window_size_matching":    (0.0, 0.5),
+    "noise_threshold":         (0.0, 0.25),
+    "wf_mz":                   (0.0, 5.0),
+    "wf_int":                  (0.0, 5.0),
+    "LET_threshold":           (0.0, 5.0),
+    "entropy_dimension":       (1.0, 3.0)
+}
+
+DEFAULT_PARAMS = {
+    "window_size_centroiding": 0.5,
+    "window_size_matching":    0.5,
+    "noise_threshold":         0.10,
+    "wf_mz":                   0.0,
+    "wf_int":                  1.0,
+    "LET_threshold":           0.0,
+    "entropy_dimension":       1.1
+}
+
+
+# ---------- Utilities ----------
+def parse_bound(s: str) -> Tuple[str, Tuple[float, float]]:
+    # "name=min:max" → (name, (min, max))
+    if "=" not in s or ":" not in s:
+        raise argparse.ArgumentTypeError(f"Bad --bound format '{s}'. Use name=min:max")
+    name, rng = s.split("=", 1)
+    lo, hi = rng.split(":", 1)
+    try:
+        lo_f, hi_f = float(lo), float(hi)
+    except ValueError as e:
+        raise argparse.ArgumentTypeError(f"Non-numeric bound in '{s}': {e}")
+    if lo_f > hi_f:
+        raise argparse.ArgumentTypeError(f"Lower bound > upper bound in '{s}'")
+    return name.strip(), (lo_f, hi_f)
+
+
+def parse_default(s: str) -> Tuple[str, float]:
+    # "name=value" → (name, value)
+    if "=" not in s:
+        raise argparse.ArgumentTypeError(f"Bad --default format '{s}'. Use name=value")
+    name, val = s.split("=", 1)
+    try:
+        v = float(val)
+    except ValueError as e:
+        raise argparse.ArgumentTypeError(f"Non-numeric default in '{s}': {e}")
+    return name.strip(), v
+
+
+def _vector_to_full_params(X: np.ndarray, default_params: Dict[str, float], optimize_params: List[str]) -> Dict[str, float]:
+    params = dict(default_params)
+    for name, val in zip(optimize_params, X):
+        params[name] = float(val)
+    return params
+
+
+# ---------- Objective wrappers (top-level, pickle-friendly) ----------
+def objective_HRMS(X: np.ndarray, ctx: dict) -> float:
+    p = _vector_to_full_params(X, ctx["default_params"], ctx["optimize_params"])
+    acc = get_acc_HRMS(
+        ctx["df_query"], ctx["df_reference"],
+        ctx["uq"], ctx["ur"],
+        ctx["similarity_measure"], ctx["weights"], ctx["spectrum_preprocessing_order"],
+        ctx["mz_min"], ctx["mz_max"], ctx["int_min"], ctx["int_max"],
+        p["window_size_centroiding"], p["window_size_matching"], p["noise_threshold"],
+        p["wf_mz"], p["wf_int"], p["LET_threshold"],
+        p["entropy_dimension"],
+        ctx["high_quality_reference_library"],
+        verbose=False
+    )
+    print(f"\n{ctx['optimize_params']} = {np.array(X)}\naccuracy: {acc*100}%")
+    return 1.0 - acc
+
+
+def objective_NRMS(X: np.ndarray, ctx: dict) -> float:
+    p = _vector_to_full_params(X, ctx["default_params"], ctx["optimize_params"])
+    acc = get_acc_NRMS(
+        ctx["df_query"], ctx["df_reference"],
+        ctx["uq"], ctx["ur"],
+        ctx["similarity_measure"], ctx["weights"], ctx["spectrum_preprocessing_order"],
+        ctx["mz_min"], ctx["mz_max"], ctx["int_min"], ctx["int_max"],
+        p["noise_threshold"], p["wf_mz"], p["wf_int"], p["LET_threshold"], p["entropy_dimension"],
+        ctx["high_quality_reference_library"],
+        verbose=False
+    )
+    print(f"\n{ctx['optimize_params']} = {np.array(X)}\naccuracy: {acc*100}%")
+    return 1.0 - acc
+
+
+# ---------- Main CLI ----------
+def main():
+    p = argparse.ArgumentParser(
+        description="Parameter tuning via Differential Evolution for HRMS/NRMS using pycompound."
+    )
+    p.add_argument("--chromatography_platform", choices=["HRMS", "NRMS"], default="HRMS", help="Chromatography Platform.")
+    p.add_argument("--query_data", required=True, help="Path to query CSV (must contain 'id' column).")
+    p.add_argument("--reference_data", required=True, nargs="+", help="Path(s) to reference CSV(s) (must contain 'id').")
+    p.add_argument("--similarity_measure", default="cosine", choices=["cosine", "renyi", "tsallis"], help="Similarity measure.")
+    p.add_argument("--weights", default="", help="Weights spec; empty means None.")
+    p.add_argument("--spectrum-order", default="CNMWL", help="Spectrum preprocessing order string.")
+    p.add_argument("--mz-min", type=float, default=0.0)
+    p.add_argument("--mz-max", type=float, default=999_999_999.0)
+    p.add_argument("--int-min", type=float, default=0.0)
+    p.add_argument("--int-max", type=float, default=999_999_999.0)
+    p.add_argument("--hq-ref-lib", action="store_true", help="Use high-quality reference library flag.")
+    p.add_argument("--opt", nargs="+", default=["window_size_centroiding", "noise_threshold", "wf_mz", "wf_int"],
+                   help=f"Parameters to optimize (subset of {ALL_PARAMS}).")
+    p.add_argument("--bound", action="append", default=[], type=parse_bound,
+                   help="Bound spec 'name=min:max'. Repeatable.")
+    p.add_argument("--default", dest="defaults", action="append", default=[], type=parse_default,
+                   help="Override a default 'name=value' for non-optimized params or initial values.")
+    p.add_argument("--maxiter", type=int, default=15)
+    p.add_argument("--seed", type=int, default=1)
+    p.add_argument("--workers", type=int, default=-1, help="Use -1 for all cores; 1 to disable parallelism.")
+    args = p.parse_args()
+
+    unknown = [x for x in args.opt if x not in ALL_PARAMS]
+    if unknown:
+        sys.exit(f"Error: unknown --opt params: {unknown}")
+
+    qpath = Path(args.query_data)
+    if not qpath.exists():
+        sys.exit(f"Query CSV not found: {qpath}")
+
+    df_query = pd.read_csv(qpath)
+    if "id" not in df_query.columns:
+        sys.exit("Query CSV must contain an 'id' column.")
+
+    ref_paths = [Path(pth) for pth in args.reference_data]
+    for r in ref_paths:
+        if not r.exists():
+            sys.exit(f"Reference CSV not found: {r}")
+    df_reference = pd.concat([pd.read_csv(r) for r in ref_paths], axis=0, ignore_index=True)
+    if "id" not in df_reference.columns:
+        sys.exit("Reference CSV must contain an 'id' column.")
+
+    uq = df_query["id"].unique().tolist()
+    ur = df_reference["id"].unique().tolist()
+
+    default_params = dict(DEFAULT_PARAMS)
+    for name, val in args.defaults:
+        if name not in DEFAULT_PARAMS:
+            sys.exit(f"--default refers to unknown parameter '{name}'. Allowed: {list(DEFAULT_PARAMS)}")
+        default_params[name] = val
+
+    param_bounds: Dict[str, Tuple[float, float]] = dict(SUGGESTED_BOUNDS)
+    for name, (lo, hi) in args.bound:
+        if name not in SUGGESTED_BOUNDS:
+            sys.exit(f"--bound refers to unknown parameter '{name}'. Allowed: {list(SUGGESTED_BOUNDS)}")
+        param_bounds[name] = (lo, hi)
+
+    bounds = [param_bounds[p] for p in args.opt]
+
+    ctx = dict(
+        df_query=df_query,
+        df_reference=df_reference,
+        uq=uq,
+        ur=ur,
+        similarity_measure=args.similarity_measure,
+        weights=(None if args.weights.strip() == "" else args.weights),
+        spectrum_preprocessing_order=args.spectrum_order,
+        mz_min=float(args.mz_min),
+        mz_max=float(args.mz_max),
+        int_min=float(args.int_min),
+        int_max=float(args.int_max),
+        high_quality_reference_library=bool(args.hq_ref_lib),
+        default_params=default_params,
+        optimize_params=args.opt,
+    )
+
+    history_acc: List[float] = []
+
+    def _cb(xk, convergence):
+        if args.chromatography_platform == "HRMS":
+            acc_pct = (1.0 - objective_HRMS(xk, ctx)) * 100.0
+        else:
+            acc_pct = (1.0 - objective_NRMS(xk, ctx)) * 100.0
+        history_acc.append(acc_pct)
+
+    objective = objective_HRMS if args.chromatography_platform == "HRMS" else objective_NRMS
+
+    result = differential_evolution(
+        objective,
+        bounds=bounds,
+        args=(ctx,),
+        maxiter=int(args.maxiter),
+        tol=0.0,
+        seed=int(args.seed),
+        workers=int(args.workers),
+        callback=_cb,
+    )
+
+    best_params = _vector_to_full_params(result.x, default_params, args.opt)
+    best_acc_pct = (1.0 - result.fun) * 100.0
+
+    print("\n=== Differential Evolution Result ===")
+    print(f"Mode: {args.chromatography_platform}")
+    print(f"Optimized over: {args.opt}")
+    print("Best values (selected params):")
+    for name in args.opt:
+        print(f"  {name}: {best_params[name]}")
+    print("\nFull parameter set used in final evaluation:")
+    for k in ALL_PARAMS:
+        print(f"  {k}: {best_params[k]}")
+    print(f"\nBest accuracy: {best_acc_pct:.3f}%")
+    print(f"DE raw: success={result.success}, nfev={result.nfev}, nit={result.nit}, message='{result.message}'")
+
+if __name__ == "__main__":
+    main()
+

pycompound-0.1.2/src/pycompound/tuning_CLI.py → pycompound-0.1.4/src/pycompound/tuning_CLI_grid.py

@@ -1,6 +1,6 @@
 
-from pycompound.spec_lib_matching import tune_params_on_HRMS_data
-from pycompound.spec_lib_matching import tune_params_on_NRMS_data
+from pycompound.spec_lib_matching import tune_params_on_HRMS_data_grid
+from pycompound.spec_lib_matching import tune_params_on_NRMS_data_grid
 import argparse
 import json
 from pathlib import Path
@@ -61,9 +61,9 @@ grid['entropy_dimension'] = [float(x) for x in grid['entropy_dimension']]
 
 
 if args.chromatography_platform == 'HRMS':
-    tune_params_on_HRMS_data(query_data=args.query_data, reference_data=args.reference_data, grid=grid, output_path=args.output_path)
+    tune_params_on_HRMS_data_grid(query_data=args.query_data, reference_data=args.reference_data, grid=grid, output_path=args.output_path)
 
 if args.chromatography_platform == 'NRMS':
-    tune_params_on_NRMS_data(query_data=args.query_data, reference_data=args.reference_data, grid=grid, output_path=args.output_path)
+    tune_params_on_NRMS_data_grid(query_data=args.query_data, reference_data=args.reference_data, grid=grid, output_path=args.output_path)
 
 

{pycompound-0.1.2 → pycompound-0.1.4/src/pycompound.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pycompound
-Version: 0.1.2
+Version: 0.1.4
 Summary: Python package to perform compound identification in mass spectrometry via spectral library matching.
 Author-email: Hunter Dlugas <fy7392@wayne.edu>
 License-Expression: MIT

{pycompound-0.1.2 → pycompound-0.1.4}/src/pycompound.egg-info/SOURCES.txt

@@ -9,7 +9,8 @@ src/pycompound/processing.py
 src/pycompound/similarity_measures.py
 src/pycompound/spec_lib_matching.py
 src/pycompound/spec_lib_matching_CLI.py
-src/pycompound/tuning_CLI.py
+src/pycompound/tuning_CLI_DE.py
+src/pycompound/tuning_CLI_grid.py
 src/pycompound.egg-info/PKG-INFO
 src/pycompound.egg-info/SOURCES.txt
 src/pycompound.egg-info/dependency_links.txt

{pycompound-0.1.2 → pycompound-0.1.4}/tests/test_spec_lib_matching.py

@@ -5,6 +5,7 @@ from pathlib import Path
 import os
 
 
+'''
 print('\n\ntest #1:')
 run_spec_lib_matching_on_HRMS_data(query_data=f'{Path.cwd()}/data/lcms_query_library.csv', reference_data=f'{Path.cwd()}/data/lcms_reference_library.csv', similarity_measure='hello')
 
@@ -34,6 +35,7 @@ run_spec_lib_matching_on_HRMS_data(query_data=f'{Path.cwd()}/data/lcms_query_lib
 
 print('\n\ntest #10:')
 run_spec_lib_matching_on_HRMS_data(query_data=f'{Path.cwd()}/data/lcms_query_library.csv', reference_data=f'{Path.cwd()}/data/lcms_reference_library.csv', normalization_method='tanh')
+'''
 
 print('\n\ntest #11:')
 run_spec_lib_matching_on_HRMS_data(query_data=f'{Path.cwd()}/data/lcms_query_library.csv', reference_data=f'{Path.cwd()}/data/lcms_reference_library.csv', similarity_measure='tsallis', wf_mz=2, wf_intensity=0.5, entropy_dimension=2, n_top_matches_to_save=3, print_id_results=True)

pycompound-0.1.4/tests/test_tuning.py

@@ -0,0 +1,52 @@
+
+from pycompound.spec_lib_matching import tune_params_on_HRMS_data_grid
+from pycompound.spec_lib_matching import tune_params_on_NRMS_data_grid
+from pycompound.spec_lib_matching import tune_params_DE
+from pathlib import Path
+import os
+
+
+print('\n\ntest #1:')
+tune_params_on_HRMS_data_grid(query_data=f'{Path.cwd()}/data/tuning/lcms_query_library.csv',
+                              reference_data=f'{Path.cwd()}/data/lcms_reference_library.csv',
+                              output_path=f'{Path.cwd()}/tuning_param_output_test1.txt')
+
+print('\n\ntest #2:')
+tune_params_on_HRMS_data_grid(query_data=f'{Path.cwd()}/data/tuning/lcms_query_library.csv',
+                              reference_data=f'{Path.cwd()}/data/lcms_reference_library.csv',
+                              grid={'similarity_measure':['cosine'], 'spectrum_preprocessing_order':['FCNMWL'], 'mz_min':[0], 'mz_max':[9999999], 'int_min':[0], 'int_max':[99999999], 'window_size_centroiding':[0.5], 'window_size_matching':[0.1,0.5], 'noise_threshold':[0.0], 'wf_mz':[0.0], 'wf_int':[1.0], 'LET_threshold':[0.0], 'entropy_dimension':[1.1], 'high_quality_reference_library':[False]},
+                              output_path=f'{Path.cwd()}/tuning_param_output_test2.txt')
+
+print('\n\ntest #3:')
+tune_params_on_NRMS_data_grid(query_data=f'{Path.cwd()}/data/tuning/gcms_query_library.csv',
+                              reference_data=f'{Path.cwd()}/data/gcms_reference_library.csv',
+                              output_path=f'{Path.cwd()}/tuning_param_output_test3.txt')
+
+print('\n\ntest #4:')
+tune_params_on_NRMS_data_grid(query_data=f'{Path.cwd()}/data/tuning/gcms_query_library.csv',
+                              reference_data=f'{Path.cwd()}/data/gcms_reference_library.csv',
+                              grid={'similarity_measure':['cosine','shannon'], 'spectrum_preprocessing_order':['FNLW'], 'mz_min':[0], 'mz_max':[9999999], 'int_min':[0], 'int_max':[99999999], 'noise_threshold':[0.0,0.1], 'wf_mz':[0.0], 'wf_int':[1.0], 'LET_threshold':[0.0,3.0], 'entropy_dimension':[1.1], 'high_quality_reference_library':[False]},
+                              output_path=f'{Path.cwd()}/tuning_param_output_test4.txt')
+
+print('\n\ntest #5:')
+tune_params_on_HRMS_data_grid(query_data=f'{Path.cwd()}/data/tuning/lcms_query_library.csv',
+                              reference_data=f'{Path.cwd()}/data/lcms_reference_library.csv',
+                              grid={'similarity_measure':['cosine'], 'weight':[{'Cosine':0.2, 'Shannon':0.2, 'Renyi':0.3, 'Tsallis':0.3},{'Cosine':0.25, 'Shannon':0.25, 'Renyi':0.25, 'Tsallis':0.25}], 'spectrum_preprocessing_order':['FCNMWL'], 'mz_min':[0], 'mz_max':[9999999], 'int_min':[0], 'int_max':[99999999], 'window_size_centroiding':[0.5], 'window_size_matching':[0.5], 'noise_threshold':[0.0], 'wf_mz':[0.0], 'wf_int':[1.0], 'LET_threshold':[0.0,3], 'entropy_dimension':[1.1], 'high_quality_reference_library':[False,True]},
+                              output_path=f'{Path.cwd()}/tuning_param_output_test5.txt')
+
+print('\n\ntest #6:')
+tune_params_DE(query_data=f'{Path.cwd()}/data/tuning/tuning_data/filtered/lcms_query_data.csv',
+               reference_data=f'{Path.cwd()}/data/tuning/tuning_data/filtered/lcms_reference_data.csv',
+               similarity_measure='shannon',
+               optimize_params=["window_size_matching","noise_threshold","wf_mz","wf_int"],
+               param_bounds={"window_size_matching":(0.0,0.5),"noise_threshold":(0.0,0.25),"wf_mz":(0.0,5.0),"wf_int":(0.0,5.0)},
+               default_params={"window_size_centroiding": 0.5, "window_size_matching":0.5, "noise_threshold":0.10, "wf_mz":0.0, "wf_int":1.0, "LET_threshold":0.0, "entropy_dimension":1.1})
+
+print('\n\ntest #7:')
+tune_params_DE(query_data=f'{Path.cwd()}/data/tuning/tuning_data/filtered/gcms_query_data.csv',
+               reference_data=f'{Path.cwd()}/data/tuning/tuning_data/filtered/gcms_reference_data.csv',
+               similarity_measure='renyi',
+               optimize_params=["wf_mz","wf_int","LET_threshold","entropy_dimension"],
+               param_bounds={"wf_mz":(0.0,5.0),"wf_int":(0.0,5.0),"LET_threshold":(0,5),"entropy_dimension":(1.01,3)},
+               default_params={"noise_threshold":0.10, "wf_mz":0.0, "wf_int":1.0, "LET_threshold":0.0, "entropy_dimension":1.1})
+

pycompound-0.1.2/tests/test_tuning.py

@@ -1,21 +0,0 @@
-
-from pycompound.spec_lib_matching import tune_params_on_HRMS_data
-from pycompound.spec_lib_matching import tune_params_on_NRMS_data
-from pathlib import Path
-import os
-
-print('\n\ntest #1:')
-tune_params_on_HRMS_data(query_data=f'{Path.cwd()}/data/tuning/lcms_query_library.csv', reference_data=f'{Path.cwd()}/data/lcms_reference_library.csv', output_path=f'{Path.cwd()}/tuning_param_output_test1.txt')
-
-print('\n\ntest #2:')
-tune_params_on_HRMS_data(query_data=f'{Path.cwd()}/data/tuning/lcms_query_library.csv', reference_data=f'{Path.cwd()}/data/lcms_reference_library.csv', grid={'similarity_measure':['cosine'], 'spectrum_preprocessing_order':['FCNMWL'], 'mz_min':[0], 'mz_max':[9999999], 'int_min':[0], 'int_max':[99999999], 'window_size_centroiding':[0.5], 'window_size_matching':[0.1,0.5], 'noise_threshold':[0.0], 'wf_mz':[0.0], 'wf_int':[1.0], 'LET_threshold':[0.0], 'entropy_dimension':[1.1], 'high_quality_reference_library':[False]}, output_path=f'{Path.cwd()}/tuning_param_output_test2.txt')
-
-print('\n\ntest #3:')
-tune_params_on_NRMS_data(query_data=f'{Path.cwd()}/data/tuning/gcms_query_library.csv', reference_data=f'{Path.cwd()}/data/gcms_reference_library.csv', output_path=f'{Path.cwd()}/tuning_param_output_test3.txt')
-
-print('\n\ntest #4:')
-tune_params_on_NRMS_data(query_data=f'{Path.cwd()}/data/tuning/gcms_query_library.csv', reference_data=f'{Path.cwd()}/data/gcms_reference_library.csv', grid={'similarity_measure':['cosine','shannon'], 'spectrum_preprocessing_order':['FNLW'], 'mz_min':[0], 'mz_max':[9999999], 'int_min':[0], 'int_max':[99999999], 'noise_threshold':[0.0,0.1], 'wf_mz':[0.0], 'wf_int':[1.0], 'LET_threshold':[0.0,3.0], 'entropy_dimension':[1.1], 'high_quality_reference_library':[False]}, output_path=f'{Path.cwd()}/tuning_param_output_test4.txt')
-
-print('\n\ntest #5:')
-tune_params_on_HRMS_data(query_data=f'{Path.cwd()}/data/tuning/lcms_query_library.csv', reference_data=f'{Path.cwd()}/data/lcms_reference_library.csv', grid={'similarity_measure':['cosine'], 'weight':[{'Cosine':0.2, 'Shannon':0.2, 'Renyi':0.3, 'Tsallis':0.3},{'Cosine':0.25, 'Shannon':0.25, 'Renyi':0.25, 'Tsallis':0.25}], 'spectrum_preprocessing_order':['FCNMWL'], 'mz_min':[0], 'mz_max':[9999999], 'int_min':[0], 'int_max':[99999999], 'window_size_centroiding':[0.5], 'window_size_matching':[0.5], 'noise_threshold':[0.0], 'wf_mz':[0.0], 'wf_int':[1.0], 'LET_threshold':[0.0,3], 'entropy_dimension':[1.1], 'high_quality_reference_library':[False,True]}, output_path=f'{Path.cwd()}/tuning_param_output_test5.txt')
-