Coverage for biobb_ml/dimensionality_reduction/pls_regression.py: 84%
89 statements
« prev ^ index » next coverage.py v7.6.1, created at 2024-10-03 14:57 +0000
1#!/usr/bin/env python3
3"""Module containing the PLS_Regression class and the command line interface."""
4import argparse
5import warnings
6import pandas as pd
7from biobb_common.generic.biobb_object import BiobbObject
8from sklearn.cross_decomposition import PLSRegression
9from sklearn.model_selection import cross_val_predict
10from sklearn.metrics import mean_squared_error, r2_score
11from biobb_common.configuration import settings
12from biobb_common.tools import file_utils as fu
13from biobb_common.tools.file_utils import launchlogger
14from biobb_ml.dimensionality_reduction.common import check_input_path, check_output_path, getHeader, getIndependentVars, getIndependentVarsList, getTarget, getTargetValue, PLSRegPlot
17class PLS_Regression(BiobbObject):
18 """
19 | biobb_ml PLS_Regression
20 | Wrapper of the scikit-learn PLSRegression method.
21 | Gives results for a Partial Least Square (PLS) Regression. Visit the `PLSRegression documentation page <https://scikit-learn.org/stable/modules/generated/sklearn.cross_decomposition.PLSRegression.html>`_ in the sklearn official website for further information.
23 Args:
24 input_dataset_path (str): Path to the input dataset. File type: input. `Sample file <https://github.com/bioexcel/biobb_ml/raw/master/biobb_ml/test/data/dimensionality_reduction/dataset_pls_regression.csv>`_. Accepted formats: csv (edam:format_3752).
25 output_results_path (str): Table with R2 and MSE for calibration and cross-validation data. File type: output. `Sample file <https://github.com/bioexcel/biobb_ml/raw/master/biobb_ml/test/reference/dimensionality_reduction/ref_output_results_pls_regression.csv>`_. Accepted formats: csv (edam:format_3752).
26 output_plot_path (str) (Optional): Path to the R2 cross-validation plot. File type: output. `Sample file <https://github.com/bioexcel/biobb_ml/raw/master/biobb_ml/test/reference/dimensionality_reduction/ref_output_plot_pls_regression.png>`_. Accepted formats: png (edam:format_3603).
27 properties (dic - Python dictionary object containing the tool parameters, not input/output files):
28 * **features** (*dict*) - ({}) Features or columns from your dataset you want to use for fitting. You can specify either a list of columns names from your input dataset, a list of columns indexes or a range of columns indexes. Formats: { "columns": ["column1", "column2"] } or { "indexes": [0, 2, 3, 10, 11, 17] } or { "range": [[0, 20], [50, 102]] }. In case of mulitple formats, the first one will be picked.
29 * **target** (*dict*) - ({}) Dependent variable you want to predict from your dataset. You can specify either a column name or a column index. Formats: { "column": "column3" } or { "index": 21 }. In case of mulitple formats, the first one will be picked.
30 * **n_components** (*int*) - (5) [1~1000|1] Maximum number of components to use by default for PLS queries.
31 * **cv** (*int*) - (10) [1~10000|1] Specify the number of folds in the cross-validation splitting strategy. Value must be betwwen 2 and number of samples in the dataset.
32 * **scale** (*bool*) - (False) Whether or not to scale the input dataset.
33 * **remove_tmp** (*bool*) - (True) [WF property] Remove temporal files.
34 * **restart** (*bool*) - (False) [WF property] Do not execute if output files exist.
35 * **sandbox_path** (*str*) - ("./") [WF property] Parent path to the sandbox directory.
37 Examples:
38 This is a use example of how to use the building block from Python::
40 from biobb_ml.dimensionality_reduction.pls_regression import pls_regression
41 prop = {
42 'features': {
43 'columns': [ 'column1', 'column2', 'column3' ]
44 },
45 'target': {
46 'column': 'target'
47 },
48 'n_components': 12,
49 'cv': 10
50 }
51 pls_regression(input_dataset_path='/path/to/myDataset.csv',
52 output_results_path='/path/to/newTable.csv',
53 output_plot_path='/path/to/newPlot.png',
54 properties=prop)
56 Info:
57 * wrapped_software:
58 * name: scikit-learn PLSRegression
59 * version: >=0.24.2
60 * license: BSD 3-Clause
61 * ontology:
62 * name: EDAM
63 * schema: http://edamontology.org/EDAM.owl
65 """
67 def __init__(self, input_dataset_path, output_results_path,
68 output_plot_path=None, properties=None, **kwargs) -> None:
69 properties = properties or {}
71 # Call parent class constructor
72 super().__init__(properties)
73 self.locals_var_dict = locals().copy()
75 # Input/Output files
76 self.io_dict = {
77 "in": {"input_dataset_path": input_dataset_path},
78 "out": {"output_results_path": output_results_path, "output_plot_path": output_plot_path}
79 }
81 # Properties specific for BB
82 self.features = properties.get('features', [])
83 self.target = properties.get('target', '')
84 self.n_components = properties.get('n_components', 5)
85 self.cv = properties.get('cv', 10)
86 self.scale = properties.get('scale', False)
87 self.properties = properties
89 # Check the properties
90 self.check_properties(properties)
91 self.check_arguments()
93 def check_data_params(self, out_log, err_log):
94 """ Checks all the input/output paths and parameters """
95 self.io_dict["in"]["input_dataset_path"] = check_input_path(self.io_dict["in"]["input_dataset_path"], "input_dataset_path", out_log, self.__class__.__name__)
96 self.io_dict["out"]["output_results_path"] = check_output_path(self.io_dict["out"]["output_results_path"], "output_results_path", False, out_log, self.__class__.__name__)
97 if self.io_dict["out"]["output_plot_path"]:
98 self.io_dict["out"]["output_plot_path"] = check_output_path(self.io_dict["out"]["output_plot_path"], "output_plot_path", True, out_log, self.__class__.__name__)
100 def warn(*args, **kwargs):
101 pass
103 @launchlogger
104 def launch(self) -> int:
105 """Execute the :class:`PLS_Regression <dimensionality_reduction.pls_regression.PLS_Regression>` dimensionality_reduction.pls_regression.PLS_Regression object."""
107 # trick for disable warnings in interations
108 warnings.warn = self.warn
110 # check input/output paths and parameters
111 self.check_data_params(self.out_log, self.err_log)
113 # Setup Biobb
114 if self.check_restart():
115 return 0
116 self.stage_files()
118 # load dataset
119 fu.log('Getting dataset from %s' % self.io_dict["in"]["input_dataset_path"], self.out_log, self.global_log)
120 if 'columns' in self.features:
121 labels = getHeader(self.io_dict["in"]["input_dataset_path"])
122 skiprows = 1
123 else:
124 labels = None
125 skiprows = None
126 data = pd.read_csv(self.io_dict["in"]["input_dataset_path"], header=None, sep="\\s+|;|:|,|\t", engine="python", skiprows=skiprows, names=labels)
128 # declare inputs, targets and weights
129 # the inputs are all the features
130 features = getIndependentVars(self.features, data, self.out_log, self.__class__.__name__)
131 fu.log('Features: [%s]' % (getIndependentVarsList(self.features)), self.out_log, self.global_log)
132 # target
133 y = getTarget(self.target, data, self.out_log, self.__class__.__name__)
134 fu.log('Target: %s' % (getTargetValue(self.target)), self.out_log, self.global_log)
136 # get rid of baseline and linear variations calculating second derivative
137 # fu.log('Performing second derivative on the data', self.out_log, self.global_log)
138 # self.window_length = getWindowLength(17, features.shape[1])
139 # X = savgol_filter(features, window_length = self.window_length, polyorder = 2, deriv = 2)
140 X = features
142 # define PLS object with optimal number of components
143 model = PLSRegression(n_components=self.n_components, scale=self.scale)
144 # fit to the entire dataset
145 model.fit(X, y)
146 y_c = model.predict(X)
147 # cross-validation
148 y_cv = cross_val_predict(model, X, y, cv=self.cv)
149 # calculate scores for calibration and cross-validation
150 score_c = r2_score(y, y_c)
151 score_cv = r2_score(y, y_cv)
152 # calculate mean squared error for calibration and cross validation
153 mse_c = mean_squared_error(y, y_c)
154 mse_cv = mean_squared_error(y, y_cv)
155 # create scores table
156 r2_table = pd.DataFrame()
157 r2_table["feature"] = ['R2 calib', 'R2 CV', 'MSE calib', 'MSE CV']
158 r2_table['coefficient'] = [score_c, score_cv, mse_c, mse_cv]
160 fu.log('Generating scores table\n\nR2 & MSE TABLE\n\n%s\n' % r2_table, self.out_log, self.global_log)
162 # save results table
163 fu.log('Saving R2 & MSE table to %s' % self.io_dict["out"]["output_results_path"], self.out_log, self.global_log)
164 r2_table.to_csv(self.io_dict["out"]["output_results_path"], index=False, header=True, float_format='%.3f')
166 # mse plot
167 if self.io_dict["out"]["output_plot_path"]:
168 fu.log('Saving MSE plot to %s' % self.io_dict["out"]["output_plot_path"], self.out_log, self.global_log)
169 plot = PLSRegPlot(y, y_c, y_cv)
170 plot.savefig(self.io_dict["out"]["output_plot_path"], dpi=150)
172 # Copy files to host
173 self.copy_to_host()
175 self.tmp_files.extend([
176 self.stage_io_dict.get("unique_dir")
177 ])
178 self.remove_tmp_files()
180 self.check_arguments(output_files_created=True, raise_exception=False)
182 return 0
185def pls_regression(input_dataset_path: str, output_results_path: str, output_plot_path: str = None, properties: dict = None, **kwargs) -> int:
186 """Execute the :class:`PLS_Regression <dimensionality_reduction.pls_regression.PLS_Regression>` class and
187 execute the :meth:`launch() <dimensionality_reduction.pls_regression.PLS_Regression.launch>` method."""
189 return PLS_Regression(input_dataset_path=input_dataset_path,
190 output_results_path=output_results_path,
191 output_plot_path=output_plot_path,
192 properties=properties, **kwargs).launch()
195def main():
196 """Command line execution of this building block. Please check the command line documentation."""
197 parser = argparse.ArgumentParser(description="Wrapper of the scikit-learn PLSRegression method.", formatter_class=lambda prog: argparse.RawTextHelpFormatter(prog, width=99999))
198 parser.add_argument('--config', required=False, help='Configuration file')
200 # Specific args of each building block
201 required_args = parser.add_argument_group('required arguments')
202 required_args.add_argument('--input_dataset_path', required=True, help='Path to the input dataset. Accepted formats: csv.')
203 required_args.add_argument('--output_results_path', required=True, help='Table with R2 and MSE for calibration and cross-validation data. Accepted formats: csv.')
204 parser.add_argument('--output_plot_path', required=False, help='Path to the R2 cross-validation plot. Accepted formats: png.')
206 args = parser.parse_args()
207 args.config = args.config or "{}"
208 properties = settings.ConfReader(config=args.config).get_prop_dic()
210 # Specific call of each building block
211 pls_regression(input_dataset_path=args.input_dataset_path,
212 output_results_path=args.output_results_path,
213 output_plot_path=args.output_plot_path,
214 properties=properties)
217if __name__ == '__main__':
218 main()