Coverage for biobb_ml/classification/classification_predict.py: 74%
96 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 ClassificationPredict class and the command line interface."""
4import argparse
5import pandas as pd
6import joblib
7from biobb_common.generic.biobb_object import BiobbObject
8from sklearn.preprocessing import StandardScaler
9from sklearn import linear_model
10from sklearn.neighbors import KNeighborsClassifier
11from sklearn.tree import DecisionTreeClassifier
12from sklearn import ensemble
13from sklearn import svm
14from biobb_common.configuration import settings
15from biobb_common.tools import file_utils as fu
16from biobb_common.tools.file_utils import launchlogger
17from biobb_ml.classification.common import check_input_path, check_output_path, getHeader, get_list_of_predictors, get_keys_of_predictors
20class ClassificationPredict(BiobbObject):
21 """
22 | biobb_ml ClassificationPredict
23 | Makes predictions from an input dataset and a given classification model.
24 | Makes predictions from an input dataset (provided either as a file or as a dictionary property) and a given classification model trained with `DecisionTreeClassifier <https://scikit-learn.org/stable/modules/generated/sklearn.tree.DecisionTreeClassifier.html>`_, `KNeighborsClassifier <https://scikit-learn.org/stable/modules/generated/sklearn.neighbors.KNeighborsClassifier.html>`_, `LogisticRegression <https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html>`_, `RandomForestClassifier <https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.RandomForestClassifier.html>`_, `Support Vector Machine <https://scikit-learn.org/stable/modules/generated/sklearn.svm.SVC.html>`_ methods.
26 Args:
27 input_model_path (str): Path to the input model. File type: input. `Sample file <https://github.com/bioexcel/biobb_ml/raw/master/biobb_ml/test/data/classification/model_classification_predict.pkl>`_. Accepted formats: pkl (edam:format_3653).
28 input_dataset_path (str) (Optional): Path to the dataset to predict. File type: input. `Sample file <https://github.com/bioexcel/biobb_ml/raw/master/biobb_ml/test/data/classification/input_classification_predict.csv>`_. Accepted formats: csv (edam:format_3752).
29 output_results_path (str): Path to the output results file. File type: output. `Sample file <https://github.com/bioexcel/biobb_ml/raw/master/biobb_ml/test/reference/classification/ref_output_classification_predict.csv>`_. Accepted formats: csv (edam:format_3752).
30 properties (dic - Python dictionary object containing the tool parameters, not input/output files):
31 * **predictions** (*list*) - (None) List of dictionaries with all values you want to predict targets. It will be taken into account only in case **input_dataset_path** is not provided. Format: [{ 'var1': 1.0, 'var2': 2.0 }, { 'var1': 4.0, 'var2': 2.7 }] for datasets with headers and [[ 1.0, 2.0 ], [ 4.0, 2.7 ]] for datasets without headers.
32 * **remove_tmp** (*bool*) - (True) [WF property] Remove temporal files.
33 * **restart** (*bool*) - (False) [WF property] Do not execute if output files exist.
34 * **sandbox_path** (*str*) - ("./") [WF property] Parent path to the sandbox directory.
36 Examples:
37 This is a use example of how to use the building block from Python::
39 from biobb_ml.classification.classification_predict import classification_predict
40 prop = {
41 'predictions': [
42 {
43 'var1': 1.0,
44 'var2': 2.0
45 },
46 {
47 'var1': 4.0,
48 'var2': 2.7
49 }
50 ]
51 }
52 classification_predict(input_model_path='/path/to/myModel.pkl',
53 output_results_path='/path/to/newPredictedResults.csv',
54 input_dataset_path='/path/to/myDataset.csv',
55 properties=prop)
57 Info:
58 * wrapped_software:
59 * name: scikit-learn
60 * version: >=0.24.2
61 * license: BSD 3-Clause
62 * ontology:
63 * name: EDAM
64 * schema: http://edamontology.org/EDAM.owl
66 """
68 def __init__(self, input_model_path, output_results_path,
69 input_dataset_path=None, properties=None, **kwargs) -> None:
70 properties = properties or {}
72 # Call parent class constructor
73 super().__init__(properties)
74 self.locals_var_dict = locals().copy()
76 # Input/Output files
77 self.io_dict = {
78 "in": {"input_model_path": input_model_path, "input_dataset_path": input_dataset_path},
79 "out": {"output_results_path": output_results_path}
80 }
82 # Properties specific for BB
83 self.predictions = properties.get('predictions', [])
84 self.properties = properties
86 # Check the properties
87 self.check_properties(properties)
88 self.check_arguments()
90 def check_data_params(self, out_log, err_log):
91 """ Checks all the input/output paths and parameters """
92 self.io_dict["in"]["input_model_path"] = check_input_path(self.io_dict["in"]["input_model_path"], "input_model_path", out_log, self.__class__.__name__)
93 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__)
94 if self.io_dict["in"]["input_dataset_path"]:
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__)
97 @launchlogger
98 def launch(self) -> int:
99 """Execute the :class:`ClassificationPredict <classification.classification_predict.ClassificationPredict>` classification.classification_predict.ClassificationPredict object."""
101 # check input/output paths and parameters
102 self.check_data_params(self.out_log, self.err_log)
104 # Setup Biobb
105 if self.check_restart():
106 return 0
107 self.stage_files()
109 fu.log('Getting model from %s' % self.io_dict["in"]["input_model_path"], self.out_log, self.global_log)
111 with open(self.io_dict["in"]["input_model_path"], "rb") as f:
112 while True:
113 try:
114 m = joblib.load(f)
115 if (isinstance(m, linear_model.LogisticRegression) or isinstance(m, KNeighborsClassifier) or isinstance(m, DecisionTreeClassifier) or isinstance(m, ensemble.RandomForestClassifier) or isinstance(m, svm.SVC)):
116 new_model = m
117 if isinstance(m, StandardScaler):
118 scaler = m
119 if isinstance(m, dict):
120 variables = m
121 except EOFError:
122 break
124 if self.io_dict["in"]["input_dataset_path"]:
125 # load dataset from input_dataset_path file
126 fu.log('Getting dataset from %s' % self.io_dict["in"]["input_dataset_path"], self.out_log, self.global_log)
127 if 'columns' in variables['independent_vars']:
128 labels = getHeader(self.io_dict["in"]["input_dataset_path"])
129 skiprows = 1
130 else:
131 labels = None
132 skiprows = None
133 new_data_table = pd.read_csv(self.io_dict["in"]["input_dataset_path"], header=None, sep="\\s+|;|:|,|\t", engine="python", skiprows=skiprows, names=labels)
134 else:
135 # load dataset from properties
136 if 'columns' in variables['independent_vars']:
137 # sorting self.properties in the correct order given by variables['independent_vars']['columns']
138 index_map = {v: i for i, v in enumerate(variables['independent_vars']['columns'])}
139 predictions = []
140 for i, pred in enumerate(self.predictions):
141 sorted_pred = sorted(pred.items(), key=lambda pair: index_map[pair[0]])
142 predictions.append(dict(sorted_pred))
143 new_data_table = pd.DataFrame(data=get_list_of_predictors(predictions), columns=get_keys_of_predictors(predictions))
144 else:
145 predictions = self.predictions
146 new_data_table = pd.DataFrame(data=predictions)
148 if variables['scale']:
149 fu.log('Scaling dataset', self.out_log, self.global_log)
150 new_data = scaler.transform(new_data_table)
151 else:
152 new_data = new_data_table
154 p = new_model.predict_proba(new_data)
156 # if headers, create target column with proper label
157 if self.io_dict["in"]["input_dataset_path"] or 'columns' in variables['independent_vars']:
158 clss = ' (' + ', '.join(str(x) for x in variables['target_values']) + ')'
159 new_data_table[variables['target']['column'] + ' ' + clss] = tuple(map(tuple, p))
160 else:
161 new_data_table[len(new_data_table.columns)] = tuple(map(tuple, p))
162 fu.log('Predicting results\n\nPREDICTION RESULTS\n\n%s\n' % new_data_table, self.out_log, self.global_log)
163 fu.log('Saving results to %s' % self.io_dict["out"]["output_results_path"], self.out_log, self.global_log)
164 new_data_table.to_csv(self.io_dict["out"]["output_results_path"], index=False, header=True, float_format='%.3f')
166 # Copy files to host
167 self.copy_to_host()
169 self.tmp_files.extend([
170 self.stage_io_dict.get("unique_dir")
171 ])
172 self.remove_tmp_files()
174 self.check_arguments(output_files_created=True, raise_exception=False)
176 return 0
179def classification_predict(input_model_path: str, output_results_path: str, input_dataset_path: str = None, properties: dict = None, **kwargs) -> int:
180 """Execute the :class:`ClassificationPredict <classification.classification_predict.ClassificationPredict>` class and
181 execute the :meth:`launch() <classification.classification_predict.ClassificationPredict.launch>` method."""
183 return ClassificationPredict(input_model_path=input_model_path,
184 output_results_path=output_results_path,
185 input_dataset_path=input_dataset_path,
186 properties=properties, **kwargs).launch()
189def main():
190 """Command line execution of this building block. Please check the command line documentation."""
191 parser = argparse.ArgumentParser(description="Makes predictions from an input dataset and a given classification model.", formatter_class=lambda prog: argparse.RawTextHelpFormatter(prog, width=99999))
192 parser.add_argument('--config', required=False, help='Configuration file')
194 # Specific args of each building block
195 required_args = parser.add_argument_group('required arguments')
196 required_args.add_argument('--input_model_path', required=True, help='Path to the input model. Accepted formats: pkl.')
197 required_args.add_argument('--output_results_path', required=True, help='Path to the output results file. Accepted formats: csv.')
198 parser.add_argument('--input_dataset_path', required=False, help='Path to the dataset to predict. Accepted formats: csv.')
200 args = parser.parse_args()
201 args.config = args.config or "{}"
202 properties = settings.ConfReader(config=args.config).get_prop_dic()
204 # Specific call of each building block
205 classification_predict(input_model_path=args.input_model_path,
206 output_results_path=args.output_results_path,
207 input_dataset_path=args.input_dataset_path,
208 properties=properties)
211if __name__ == '__main__':
212 main()