CoreMS 4.0.0__py3-none-any.whl

corems/encapsulation/output/parameter_to_dict.py

@@ -0,0 +1,142 @@
+from corems.encapsulation.factory.parameters import (
+    MSParameters,
+    GCMSParameters,
+    LCMSParameters
+    )
+
+
+def get_dict_all_default_data():
+    """Return a dictionary with all default parameters for MS and GCMS"""
+    ms_params = MSParameters(use_defaults=True)
+    gcms_params = GCMSParameters(use_defaults=True)
+
+    return {
+        "MolecularFormulaSearch": ms_params.molecular_search.__dict__,
+        "Transient": ms_params.transient.__dict__,
+        "MassSpectrum": ms_params.mass_spectrum.__dict__,
+        "MassSpecPeak": ms_params.ms_peak.__dict__,
+        "DataInput": ms_params.data_input.__dict__,
+        "MolecularSearch": gcms_params.molecular_search.__dict__,
+        "GasChromatograph": gcms_params.gc_ms.__dict__,
+    }
+
+
+def get_dict_data_lcms(lcms_obj):
+    """Return a dictionary with all parameters for LCMSBase object
+
+    Parameters
+    ----------
+    lcms_obj: LCMSBase
+        LCMSBase object
+
+    Returns
+    -------
+    dict
+        dictionary with all parameters for LCMSBase object
+    """
+    output_dict = {}
+    output_dict["LiquidChromatograph"] = lcms_obj.parameters.lc_ms.__dict__
+    output_dict["mass_spectrum"] = {}
+    for key, value in lcms_obj.parameters.mass_spectrum.items():
+        output_dict["mass_spectrum"][key] = {}
+        for k, v in value.__dict__.items():
+            output_dict["mass_spectrum"][key][k] = v.__dict__
+    return output_dict
+
+
+def get_dict_lcms_default_data():
+    """Return a dictionary with all default parameters for LCMS"""
+    default_params = LCMSParameters(use_defaults=True)
+
+    output_dict = {}
+    output_dict["LiquidChromatograph"] = default_params.lc_ms.__dict__
+    output_dict["mass_spectrum"] = {}
+    for key, value in default_params.mass_spectrum.items():
+        output_dict["mass_spectrum"][key] = {}
+        for k, v in value.__dict__.items():
+            output_dict["mass_spectrum"][key][k] = v.__dict__
+    return output_dict
+
+
+def get_dict_data_ms(mass_spec):
+    """Return a dictionary with all parameters for MassSpectrum object
+
+    Parameters
+    ----------
+    mass_spec: MassSpectrum
+        MassSpectrum object
+
+    Returns
+    -------
+    dict
+        dictionary with all parameters for MassSpectrum object
+    """
+    ms_params = mass_spec.parameters
+    return {
+        "MolecularFormulaSearch": ms_params.molecular_search.__dict__,
+        "Transient": ms_params.transient.__dict__,
+        "MassSpectrum": ms_params.mass_spectrum.__dict__,
+        "MassSpecPeak": ms_params.ms_peak.__dict__,
+        "DataInput": ms_params.data_input.__dict__,
+    }
+
+
+def get_dict_ms_default_data():
+    """Return a dictionary with all default parameters for MS including data input"""
+    ms_params = MSParameters(use_defaults=True)
+
+    return {
+        "MolecularFormulaSearch": ms_params.molecular_search.__dict__,
+        "Transient": ms_params.transient.__dict__,
+        "MassSpectrum": ms_params.mass_spectrum.__dict__,
+        "MassSpecPeak": ms_params.ms_peak.__dict__,
+        "DataInput": ms_params.data_input.__dict__,
+    }
+
+
+def get_dict_gcms_default_data():
+    """Return a dictionary with all default parameters for GCMS"""
+    default_gcms_params = GCMSParameters(use_defaults=True)
+
+    return {
+        "MolecularSearch": default_gcms_params.molecular_search.__dict__,
+        "GasChromatograph": default_gcms_params.gc_ms.__dict__,
+    }
+
+
+def get_dict_data_gcms(gcms):
+    """Return a dictionary with all parameters for GCMS"""
+
+    return {
+        "MolecularSearch": gcms.molecular_search_settings.__dict__,
+        "GasChromatograph": gcms.chromatogram_settings.__dict__,
+    }
+
+
+def get_dict_data_lcms_collection(lcms_collection):
+    """Return a dictionary with all parameters for LCMSCollection object
+
+    Parameters
+    ----------
+    lcms_collection: LCMSCollection
+        LCMSCollection object
+
+    Returns
+    -------
+    dict
+        dictionary with all parameters for LCMSCollection object
+    """
+    output_dict = {}
+    output_dict["LCMSCollection"] = lcms_collection.parameters.lcms_collection.__dict__
+    return output_dict
+
+
+def get_dict_lcms_collection_default_data():
+    """Return a dictionary with all default parameters for LCMS Collection"""
+    from corems.encapsulation.factory.processingSetting import LCMSCollectionSettings
+    
+    default_params = LCMSCollectionSettings()
+
+    output_dict = {}
+    output_dict["LCMSCollection"] = default_params.__dict__
+    return output_dict

corems/encapsulation/output/parameter_to_json.py

@@ -0,0 +1,332 @@
+import json
+
+import toml
+from pathlib import Path
+
+from corems.encapsulation.output import parameter_to_dict
+from corems.encapsulation.output.parameter_to_dict import get_dict_data_lcms
+
+
+def dump_all_settings_json(filename="SettingsCoreMS.json", file_path=None):
+    """
+    Write JSON file into current directory with all the default settings for the CoreMS package.
+
+    Parameters:
+    ----------
+    filename : str, optional
+        The name of the JSON file to be created. Default is 'SettingsCoreMS.json'.
+    file_path : str or Path, optional
+        The path where the JSON file will be saved. If not provided, the file will be saved in the current working directory.
+    """
+
+    data_dict_all = parameter_to_dict.get_dict_all_default_data()
+
+    if not file_path:
+        file_path = Path.cwd() / filename
+
+    with open(
+        file_path,
+        "w",
+        encoding="utf8",
+    ) as outfile:
+        import re
+
+        # pretty print
+        output = json.dumps(
+            data_dict_all, sort_keys=False, indent=4, separators=(",", ": ")
+        )
+        output = re.sub(r'",\s+', '", ', output)
+
+        outfile.write(output)
+
+
+def dump_ms_settings_json(filename="SettingsCoreMS.json", file_path=None):
+    """
+    Write JSON file into current directory with all the mass spectrum default settings for the CoreMS package.
+
+    Parameters
+    ----------
+    filename : str, optional
+        The name of the JSON file to be created. Default is 'SettingsCoreMS.json'.
+    file_path : str or Path, optional
+        The path where the JSON file will be saved. If not provided, the file will be saved in the current working directory.
+
+    """
+    data_dict = parameter_to_dict.get_dict_ms_default_data()
+    if not file_path:
+        file_path = Path.cwd() / filename
+
+    with open(
+        file_path,
+        "w",
+        encoding="utf8",
+    ) as outfile:
+        import re
+
+        # pretty print
+        output = json.dumps(
+            data_dict, sort_keys=False, indent=4, separators=(",", ": ")
+        )
+        output = re.sub(r'",\s+', '", ', output)
+
+        outfile.write(output)
+
+
+def dump_gcms_settings_json(filename="SettingsCoreMS.json", file_path=None):
+    """
+    Write JSON file into current directory containing the default GCMS settings data.
+
+    Parameters
+    ----------
+    filename : str, optional
+        The name of the JSON file to be created. Default is 'SettingsCoreMS.json'.
+    file_path : str or Path-like object, optional
+        The path where the JSON file will be saved. If not provided, the file will be saved in the current working directory.
+    """
+
+    from pathlib import Path
+    import json
+
+    data_dict = parameter_to_dict.get_dict_gcms_default_data()
+
+    if not file_path:
+        file_path = Path.cwd() / filename
+
+    with open(
+        file_path,
+        "w",
+        encoding="utf8",
+    ) as outfile:
+        import re
+
+        # pretty print
+        output = json.dumps(
+            data_dict, sort_keys=False, indent=4, separators=(",", ": ")
+        )
+        output = re.sub(r'",\s+', '", ', output)
+
+        outfile.write(output)
+
+
+def dump_all_settings_toml(filename="SettingsCoreMS.toml", file_path=None):
+    """
+    Write TOML file into the specified file path or the current directory with all the default settings for the CoreMS package.
+
+    Parameters
+    ----------
+    filename : str, optional
+        The name of the TOML file. Defaults to 'SettingsCoreMS.toml'.
+    file_path : str or Path, optional
+        The path where the TOML file will be saved. If not provided, the file will be saved in the current directory.
+
+    """
+    from pathlib import Path
+
+    data_dict_all = parameter_to_dict.get_dict_all_default_data()
+
+    if not file_path:
+        file_path = Path.cwd() / filename
+
+    with open(
+        file_path,
+        "w",
+        encoding="utf8",
+    ) as outfile:
+        import re
+
+        output = toml.dumps(data_dict_all)
+        outfile.write(output)
+
+
+def dump_ms_settings_toml(filename="SettingsCoreMS.toml", file_path=None):
+    """
+    Write TOML file into the current directory with all the mass spectrum default settings for the CoreMS package.
+
+    Parameters
+    ----------
+    filename : str, optional
+        The name of the TOML file to be created. Default is 'SettingsCoreMS.toml'.
+    file_path : str or Path, optional
+        The path where the TOML file should be saved. If not provided, the file will be saved in the current working directory.
+
+    """
+    data_dict = parameter_to_dict.get_dict_ms_default_data()
+
+    if not file_path:
+        file_path = Path.cwd() / filename
+
+    with open(
+        file_path,
+        "w",
+        encoding="utf8",
+    ) as outfile:
+        import re
+
+        # pretty print
+        output = toml.dumps(data_dict)
+        outfile.write(output)
+
+
+def dump_gcms_settings_toml(filename="SettingsCoreMS.toml", file_path=None):
+    """
+    Write TOML file into current directory containing the default GCMS settings data.
+
+    Parameters
+    ----------
+    filename : str, optional
+        The name of the TOML file. Defaults to 'SettingsCoreMS.toml'.
+    file_path : str or Path, optional
+        The path where the TOML file will be saved. If not provided, the file will be saved in the current working directory.
+
+    """
+
+    data_dict = parameter_to_dict.get_dict_gcms_default_data()
+
+    if not file_path:
+        file_path = Path.cwd() / filename
+
+    with open(
+        file_path,
+        "w",
+        encoding="utf8",
+    ) as outfile:
+        output = toml.dumps(data_dict)
+        outfile.write(output)
+
+
+def dump_lcms_settings_json(
+    filename="SettingsCoreMS.json", file_path=None, lcms_obj=None
+):
+    """
+    Write JSON file into current directory with all the LCMS settings data for the CoreMS package.
+
+    Parameters
+    ----------
+    filename : str, optional
+        The name of the JSON file. Defaults to 'SettingsCoreMS.json'.
+    file_path : str or Path, optional
+        The path where the JSON file will be saved. If not provided, the file will be saved in the current working directory.
+    lcms_obj : object, optional
+        The LCMS object containing the settings data. If not provided, the settings data will be retrieved from the default settings.
+
+    """
+
+    if lcms_obj is None:
+        data_dict = parameter_to_dict.get_dict_lcms_default_data()
+    else:
+        data_dict = get_dict_data_lcms(lcms_obj)
+
+    if not file_path:
+        file_path = Path.cwd() / filename
+
+    with open(
+        file_path,
+        "w",
+        encoding="utf8",
+    ) as outfile:
+        outfile.write(json.dumps(data_dict, indent=4))
+
+
+def dump_lcms_settings_toml(
+    filename="SettingsCoreMS.toml", file_path=None, lcms_obj=None
+):
+    """
+    Write TOML file into current directory with all the LCMS settings data for the CoreMS package.
+
+    Parameters
+    ----------
+    filename : str, optional
+        The name of the TOML file. Defaults to 'SettingsCoreMS.toml'.
+    file_path : str or Path, optional
+        The path where the TOML file will be saved. If not provided, the file will be saved in the current working directory.
+    lcms_obj : object, optional
+        The LCMS object containing the settings data. If not provided, the settings data will be retrieved from the default settings.
+
+    """
+
+    if lcms_obj is None:
+        data_dict = parameter_to_dict.get_dict_lcms_default_data()
+    else:
+        data_dict = get_dict_data_lcms(lcms_obj)
+
+    if not file_path:
+        file_path = Path.cwd() / filename
+
+    with open(
+        file_path,
+        "w",
+        encoding="utf8",
+    ) as outfile:
+        output = toml.dumps(data_dict)
+        outfile.write(output)
+
+
+def dump_lcms_collection_settings_json(
+    filename="SettingsCoreMS.json", file_path=None, lcms_collection=None
+):
+    """Write JSON file with LCMS collection settings data.
+
+    Parameters
+    ----------
+    filename : str, optional
+        The name of the JSON file. Defaults to 'SettingsCoreMS.json'.
+    file_path : str or Path, optional
+        The path where the JSON file will be saved. If not provided, the file will be saved in the current working directory.
+    lcms_collection : LCMSCollection, optional
+        The LCMS collection object containing the settings data. If not provided, the settings data will be retrieved from the default settings.
+    """
+    from corems.encapsulation.output.parameter_to_dict import (
+        get_dict_data_lcms_collection,
+        get_dict_lcms_collection_default_data,
+    )
+
+    if lcms_collection is None:
+        data_dict = get_dict_lcms_collection_default_data()
+    else:
+        data_dict = get_dict_data_lcms_collection(lcms_collection)
+
+    if not file_path:
+        file_path = Path.cwd() / filename
+
+    with open(
+        file_path,
+        "w",
+        encoding="utf8",
+    ) as outfile:
+        outfile.write(json.dumps(data_dict, indent=4))
+
+
+def dump_lcms_collection_settings_toml(
+    filename="SettingsCoreMS.toml", file_path=None, lcms_collection=None
+):
+    """Write TOML file with LCMS collection settings data.
+
+    Parameters
+    ----------
+    filename : str, optional
+        The name of the TOML file. Defaults to 'SettingsCoreMS.toml'.
+    file_path : str or Path, optional
+        The path where the TOML file will be saved. If not provided, the file will be saved in the current working directory.
+    lcms_collection : LCMSCollection, optional
+        The LCMS collection object containing the settings data. If not provided, the settings data will be retrieved from the default settings.
+    """
+    from corems.encapsulation.output.parameter_to_dict import (
+        get_dict_data_lcms_collection,
+        get_dict_lcms_collection_default_data,
+    )
+
+    if lcms_collection is None:
+        data_dict = get_dict_lcms_collection_default_data()
+    else:
+        data_dict = get_dict_data_lcms_collection(lcms_collection)
+
+    if not file_path:
+        file_path = Path.cwd() / filename
+
+    with open(
+        file_path,
+        "w",
+        encoding="utf8",
+    ) as outfile:
+        output = toml.dumps(data_dict)
+        outfile.write(output)

corems/mass_spectra/calc/GC_Calc.py

@@ -0,0 +1,84 @@
+__author__ = "Yuri E. Corilo"
+__date__ = "Feb 13, 2020"
+
+from pathlib import Path
+import numpy as np
+from pandas import Series
+
+from corems.mass_spectra.calc import SignalProcessing as sp
+
+
+class GC_Calculations:
+    def calibrate_ri(self, ref_dict, cal_file_path):
+        if not self:
+            self.process_chromatogram()
+
+        for gcms_peak in self:
+            gcms_peak.calc_ri(ref_dict)
+
+        self.ri_pairs_ref = ref_dict
+        if isinstance(cal_file_path, str):
+            # if obj is a string it defaults to create a Path obj, pass the S3Path if needed
+            self.cal_file_path = Path(cal_file_path)
+        else:
+            self.cal_file_path = cal_file_path
+
+    def smooth_tic(self, tic):
+        implemented_smooth_method = self.chromatogram_settings.implemented_smooth_method
+
+        pol_order = self.chromatogram_settings.savgol_pol_order
+
+        window_len = self.chromatogram_settings.smooth_window
+
+        window = self.chromatogram_settings.smooth_method
+
+        return sp.smooth_signal(
+            tic, window_len, window, pol_order, implemented_smooth_method
+        )
+
+    def centroid_detector(self, tic, rt):
+        noise_std = self.chromatogram_settings.std_noise_threshold
+
+        method = self.chromatogram_settings.noise_threshold_method
+
+        # peak picking
+        min_height = self.chromatogram_settings.peak_height_min_percent
+        min_datapoints = self.chromatogram_settings.min_peak_datapoints
+
+        # baseline detection
+        max_prominence = self.chromatogram_settings.peak_max_prominence_percent
+        max_height = self.chromatogram_settings.peak_height_max_percent
+
+        peak_indexes_generator = sp.peak_detector_generator(
+            tic,
+            noise_std,
+            method,
+            rt,
+            max_height,
+            min_height,
+            max_prominence,
+            min_datapoints,
+        )
+
+        return peak_indexes_generator
+
+    def remove_outliers(self, data):
+        from numpy import percentile
+
+        q25, q75 = percentile(data, 25), percentile(data, 75)
+        iqr = q75 - q25
+        if self.parameters.verbose_processing:
+            print("Percentiles: 25th=%.3f, 75th=%.3f, IQR=%.3f" % (q25, q75, iqr))
+        # calculate the outlier cutoff
+        cut_off = iqr * 1.5
+        lower, upper = q25 - cut_off, q75 + cut_off
+        # identify outliers
+        outliers = [x for x in data if x < lower or x > upper]
+        if self.parameters.verbose_processing:
+            print("Identified outliers: %d" % len(outliers))
+        # remove outliers
+        nanfilled_outliers = Series(
+            [x if lower <= x <= upper else np.nan for x in data]
+        )
+
+        return nanfilled_outliers