annotate ensemble.xml @ 11:f4d8a82e167c draft

planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit 64158f357e708f0b60d2669d92d614f7aee34c0e
author bgruening
date Wed, 06 Jun 2018 17:42:54 -0400
parents cd595710f0c0
children a78f96f64939
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
0
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
1 <tool id="sklearn_ensemble" name="Ensemble methods" version="@VERSION@">
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
2 <description>for classification and regression</description>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
3 <macros>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
4 <import>main_macros.xml</import>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
5 </macros>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
6 <expand macro="python_requirements"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
7 <expand macro="macro_stdio"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
8 <version_command>echo "@VERSION@"</version_command>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
9 <command><![CDATA[
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
10 python "$ensemble_script" '$inputs'
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
11 ]]>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
12 </command>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
13 <configfiles>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
14 <inputs name="inputs"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
15 <configfile name="ensemble_script">
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
16 <![CDATA[
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
17 import sys
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
18 import json
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
19 import numpy as np
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
20 import sklearn.ensemble
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
21 import pandas
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
22 import pickle
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
23 from scipy.io import mmread
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
24
4
0431274c367d planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit cea052cf3b8dd4f3620253bd222e126de32e7466
bgruening
parents: 2
diff changeset
25 @COLUMNS_FUNCTION@
0431274c367d planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit cea052cf3b8dd4f3620253bd222e126de32e7466
bgruening
parents: 2
diff changeset
26
0
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
27 input_json_path = sys.argv[1]
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
28 params = json.load(open(input_json_path, "r"))
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
29
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
30 #if $selected_tasks.selected_task == "train":
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
31
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
32 algorithm = params["selected_tasks"]["selected_algorithms"]["selected_algorithm"]
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
33 options = params["selected_tasks"]["selected_algorithms"]["options"]
5
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
34 if "select_max_features" in options:
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
35 if options["select_max_features"]["max_features"] == "number_input":
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
36 options["select_max_features"]["max_features"] = options["select_max_features"]["num_max_features"]
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
37 options["select_max_features"].pop("num_max_features")
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
38 options["max_features"] = options["select_max_features"]["max_features"]
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
39 options.pop("select_max_features")
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
40 if "presort" in options:
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
41 if options["presort"] == "true":
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
42 options["presort"] = True
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
43 if options["presort"] == "false":
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
44 options["presort"] = False
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
45 if "min_samples_leaf" in options and options["min_samples_leaf"] == 1.0:
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
46 options["min_samples_leaf"] = 1
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
47 if "min_samples_split" in options and options["min_samples_split"] > 1.0:
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
48 options["min_samples_split"] = int(options["min_samples_split"])
0
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
49 input_type = params["selected_tasks"]["selected_algorithms"]["input_options"]["selected_input"]
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
50 if input_type=="tabular":
4
0431274c367d planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit cea052cf3b8dd4f3620253bd222e126de32e7466
bgruening
parents: 2
diff changeset
51 header = 'infer' if params["selected_tasks"]["selected_algorithms"]["input_options"]["header1"] else None
10
cd595710f0c0 planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit 97c4f22cdcfa6cddeeffc7b102c418a7ff12a888
bgruening
parents: 5
diff changeset
52 column_option = params["selected_tasks"]["selected_algorithms"]["input_options"]["column_selector_options_1"]["selected_column_selector_option"]
11
f4d8a82e167c planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit 64158f357e708f0b60d2669d92d614f7aee34c0e
bgruening
parents: 10
diff changeset
53 if column_option in ["by_index_number", "all_but_by_index_number"]:
10
cd595710f0c0 planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit 97c4f22cdcfa6cddeeffc7b102c418a7ff12a888
bgruening
parents: 5
diff changeset
54 c = params["selected_tasks"]["selected_algorithms"]["input_options"]["column_selector_options_1"]["col1"]
cd595710f0c0 planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit 97c4f22cdcfa6cddeeffc7b102c418a7ff12a888
bgruening
parents: 5
diff changeset
55 else:
cd595710f0c0 planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit 97c4f22cdcfa6cddeeffc7b102c418a7ff12a888
bgruening
parents: 5
diff changeset
56 c = None
2
6e6726be0728 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 641ac64ded23fbb6fe85d5f13926da12dcce4e76
bgruening
parents: 0
diff changeset
57 X = read_columns(
6e6726be0728 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 641ac64ded23fbb6fe85d5f13926da12dcce4e76
bgruening
parents: 0
diff changeset
58 "$selected_tasks.selected_algorithms.input_options.infile1",
10
cd595710f0c0 planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit 97c4f22cdcfa6cddeeffc7b102c418a7ff12a888
bgruening
parents: 5
diff changeset
59 c = c,
cd595710f0c0 planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit 97c4f22cdcfa6cddeeffc7b102c418a7ff12a888
bgruening
parents: 5
diff changeset
60 c_option = column_option,
2
6e6726be0728 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 641ac64ded23fbb6fe85d5f13926da12dcce4e76
bgruening
parents: 0
diff changeset
61 sep='\t',
4
0431274c367d planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit cea052cf3b8dd4f3620253bd222e126de32e7466
bgruening
parents: 2
diff changeset
62 header=header,
2
6e6726be0728 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 641ac64ded23fbb6fe85d5f13926da12dcce4e76
bgruening
parents: 0
diff changeset
63 parse_dates=True
6e6726be0728 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 641ac64ded23fbb6fe85d5f13926da12dcce4e76
bgruening
parents: 0
diff changeset
64 )
0
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
65 else:
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
66 X = mmread(open("$selected_tasks.selected_algorithms.input_options.infile1", 'r'))
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
67
4
0431274c367d planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit cea052cf3b8dd4f3620253bd222e126de32e7466
bgruening
parents: 2
diff changeset
68 header = 'infer' if params["selected_tasks"]["selected_algorithms"]["input_options"]["header2"] else None
10
cd595710f0c0 planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit 97c4f22cdcfa6cddeeffc7b102c418a7ff12a888
bgruening
parents: 5
diff changeset
69 column_option = params["selected_tasks"]["selected_algorithms"]["input_options"]["column_selector_options_2"]["selected_column_selector_option2"]
11
f4d8a82e167c planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit 64158f357e708f0b60d2669d92d614f7aee34c0e
bgruening
parents: 10
diff changeset
70 if column_option in ["by_index_number", "all_but_by_index_number"]:
10
cd595710f0c0 planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit 97c4f22cdcfa6cddeeffc7b102c418a7ff12a888
bgruening
parents: 5
diff changeset
71 c = params["selected_tasks"]["selected_algorithms"]["input_options"]["column_selector_options_2"]["col2"]
cd595710f0c0 planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit 97c4f22cdcfa6cddeeffc7b102c418a7ff12a888
bgruening
parents: 5
diff changeset
72 else:
cd595710f0c0 planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit 97c4f22cdcfa6cddeeffc7b102c418a7ff12a888
bgruening
parents: 5
diff changeset
73 c = None
2
6e6726be0728 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 641ac64ded23fbb6fe85d5f13926da12dcce4e76
bgruening
parents: 0
diff changeset
74 y = read_columns(
6e6726be0728 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 641ac64ded23fbb6fe85d5f13926da12dcce4e76
bgruening
parents: 0
diff changeset
75 "$selected_tasks.selected_algorithms.input_options.infile2",
10
cd595710f0c0 planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit 97c4f22cdcfa6cddeeffc7b102c418a7ff12a888
bgruening
parents: 5
diff changeset
76 c = c,
cd595710f0c0 planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit 97c4f22cdcfa6cddeeffc7b102c418a7ff12a888
bgruening
parents: 5
diff changeset
77 c_option = column_option,
2
6e6726be0728 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 641ac64ded23fbb6fe85d5f13926da12dcce4e76
bgruening
parents: 0
diff changeset
78 sep='\t',
4
0431274c367d planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit cea052cf3b8dd4f3620253bd222e126de32e7466
bgruening
parents: 2
diff changeset
79 header=header,
2
6e6726be0728 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 641ac64ded23fbb6fe85d5f13926da12dcce4e76
bgruening
parents: 0
diff changeset
80 parse_dates=True
6e6726be0728 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 641ac64ded23fbb6fe85d5f13926da12dcce4e76
bgruening
parents: 0
diff changeset
81 )
5
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
82 y=y.ravel()
0
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
83
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
84 my_class = getattr(sklearn.ensemble, algorithm)
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
85 estimator = my_class(**options)
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
86 estimator.fit(X,y)
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
87 pickle.dump(estimator,open("$outfile_fit", 'w+'), pickle.HIGHEST_PROTOCOL)
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
88
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
89 #else:
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
90 classifier_object = pickle.load(open("$selected_tasks.infile_model", 'r'))
5
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
91 header = 'infer' if params["selected_tasks"]["header"] else None
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
92 data = pandas.read_csv("$selected_tasks.infile_data", sep='\t', header=header, index_col=None, parse_dates=True, encoding=None, tupleize_cols=False)
0
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
93 prediction = classifier_object.predict(data)
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
94 prediction_df = pandas.DataFrame(prediction)
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
95 res = pandas.concat([data, prediction_df], axis=1)
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
96 res.to_csv(path_or_buf = "$outfile_predict", sep="\t", index=False)
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
97 #end if
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
98
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
99 ]]>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
100 </configfile>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
101 </configfiles>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
102 <inputs>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
103 <expand macro="sl_Conditional" model="zip">
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
104 <param name="selected_algorithm" type="select" label="Select an ensemble method:">
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
105 <option value="RandomForestClassifier" selected="true">Random forest classifier</option>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
106 <option value="AdaBoostClassifier">Ada boost classifier</option>
5
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
107 <option value="GradientBoostingClassifier">Gradient Boosting Classifier</option>
0
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
108 <option value="RandomForestRegressor">Random forest regressor</option>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
109 <option value="AdaBoostRegressor">Ada boost regressor</option>
5
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
110 <option value="GradientBoostingRegressor">Gradient Boosting Regressor</option>
0
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
111 </param>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
112 <when value="RandomForestClassifier">
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
113 <expand macro="sl_mixed_input"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
114 <section name="options" title="Advanced Options" expanded="False">
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
115 <expand macro="n_estimators"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
116 <expand macro="criterion"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
117 <expand macro="max_features"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
118 <expand macro="max_depth"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
119 <expand macro="min_samples_split"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
120 <expand macro="min_samples_leaf"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
121 <expand macro="min_weight_fraction_leaf"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
122 <expand macro="max_leaf_nodes"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
123 <expand macro="bootstrap"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
124 <expand macro="warm_start" checked="false"/>
5
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
125 <expand macro="n_jobs"/>
0
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
126 <expand macro="random_state"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
127 <expand macro="oob_score"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
128 <!--class_weight=None-->
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
129 </section>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
130 </when>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
131 <when value="AdaBoostClassifier">
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
132 <expand macro="sl_mixed_input"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
133 <section name="options" title="Advanced Options" expanded="False">
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
134 <!--base_estimator=None-->
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
135 <expand macro="n_estimators" default_value="50"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
136 <expand macro="learning_rate"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
137 <param argument="algorithm" type="select" label="Boosting algorithm" help=" ">
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
138 <option value="SAMME.R" selected="true">SAMME.R</option>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
139 <option value="SAMME">SAMME</option>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
140 </param>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
141 <expand macro="random_state"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
142 </section>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
143 </when>
5
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
144 <when value="GradientBoostingClassifier">
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
145 <expand macro="sl_mixed_input"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
146 <section name="options" title="Advanced Options" expanded="False">
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
147 <!--base_estimator=None-->
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
148 <param argument="loss" type="select" label="Loss function">
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
149 <option value="deviance" selected="true">deviance - logistic regression with probabilistic outputs</option>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
150 <option value="exponential">exponential - gradient boosting recovers the AdaBoost algorithm</option>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
151 </param>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
152 <expand macro="learning_rate" default_value='0.1'/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
153 <expand macro="n_estimators" default_value="100" help="The number of boosting stages to perform"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
154 <expand macro="max_depth" default_value="3" help="maximum depth of the individual regression estimators"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
155 <expand macro="criterion2">
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
156 <option value="friedman_mse" selected="true">friedman_mse - mean squared error with improvement score by Friedman</option>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
157 </expand>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
158 <expand macro="min_samples_split" type="float"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
159 <expand macro="min_samples_leaf" type="float" label="The minimum number of samples required to be at a leaf node"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
160 <expand macro="min_weight_fraction_leaf"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
161 <expand macro="subsample"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
162 <expand macro="max_features"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
163 <expand macro="max_leaf_nodes"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
164 <expand macro="min_impurity_decrease"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
165 <expand macro="verbose"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
166 <expand macro="warm_start" checked="false"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
167 <expand macro="random_state"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
168 <expand macro="presort"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
169 </section>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
170 </when>
0
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
171 <when value="RandomForestRegressor">
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
172 <expand macro="sl_mixed_input"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
173 <section name="options" title="Advanced Options" expanded="False">
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
174 <expand macro="n_estimators"/>
5
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
175 <expand macro="criterion2"/>
0
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
176 <expand macro="max_features"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
177 <expand macro="max_depth"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
178 <expand macro="min_samples_split"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
179 <expand macro="min_samples_leaf"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
180 <expand macro="min_weight_fraction_leaf"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
181 <expand macro="max_leaf_nodes"/>
5
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
182 <expand macro="min_impurity_decrease"/>
0
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
183 <expand macro="bootstrap"/>
5
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
184 <expand macro="oob_score"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
185 <expand macro="n_jobs"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
186 <expand macro="random_state"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
187 <expand macro="verbose"/>
0
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
188 <expand macro="warm_start" checked="false"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
189 </section>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
190 </when>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
191 <when value="AdaBoostRegressor">
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
192 <expand macro="sl_mixed_input"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
193 <section name="options" title="Advanced Options" expanded="False">
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
194 <!--base_estimator=None-->
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
195 <expand macro="n_estimators" default_value="50"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
196 <expand macro="learning_rate"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
197 <param argument="loss" type="select" label="Loss function" optional="true" help="Used when updating the weights after each boosting iteration. ">
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
198 <option value="linear" selected="true">linear</option>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
199 <option value="square">square</option>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
200 <option value="exponential">exponential</option>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
201 </param>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
202 <expand macro="random_state"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
203 </section>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
204 </when>
5
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
205 <when value="GradientBoostingRegressor">
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
206 <expand macro="sl_mixed_input"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
207 <section name="options" title="Advanced Options" expanded="False">
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
208 <param argument="loss" type="select" label="Loss function">
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
209 <option value="ls" selected="true">ls - least squares regression</option>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
210 <option value="lad">lad - least absolute deviation</option>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
211 <option value="huber">huber - combination of least squares regression and least absolute deviation</option>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
212 <option value="quantile">quantile - use alpha to specify the quantile</option>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
213 </param>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
214 <expand macro="learning_rate" default_value="0.1"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
215 <expand macro="n_estimators" default_value="100" help="The number of boosting stages to perform"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
216 <expand macro="max_depth" default_value="3" help="maximum depth of the individual regression estimators"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
217 <expand macro="criterion2">
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
218 <option value="friedman_mse" selected="true">friedman_mse - mean squared error with improvement score by Friedman</option>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
219 </expand>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
220 <expand macro="min_samples_split" type="float"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
221 <expand macro="min_samples_leaf" type="float" label="The minimum number of samples required to be at a leaf node"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
222 <expand macro="min_weight_fraction_leaf"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
223 <expand macro="subsample"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
224 <expand macro="max_features"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
225 <expand macro="max_leaf_nodes"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
226 <expand macro="min_impurity_decrease"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
227 <param argument="alpha" type="float" value="0.9" label="alpha" help="The alpha-quantile of the huber loss function and the quantile loss function" />
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
228 <!--base_estimator=None-->
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
229 <expand macro="verbose"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
230 <expand macro="warm_start" checked="false"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
231 <expand macro="random_state"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
232 <expand macro="presort"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
233 </section>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
234 </when>
0
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
235 </expand>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
236 </inputs>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
237
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
238 <expand macro="output"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
239
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
240 <tests>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
241 <test>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
242 <param name="infile1" value="train.tabular" ftype="tabular"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
243 <param name="infile2" value="train.tabular" ftype="tabular"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
244 <param name="col1" value="1,2,3,4"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
245 <param name="col2" value="5"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
246 <param name="selected_task" value="train"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
247 <param name="selected_algorithm" value="RandomForestClassifier"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
248 <param name="random_state" value="10"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
249 <output name="outfile_fit" file="rfc_model01" compare="sim_size" delta="500"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
250 </test>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
251 <test>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
252 <param name="infile_model" value="rfc_model01" ftype="zip"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
253 <param name="infile_data" value="test.tabular" ftype="tabular"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
254 <param name="selected_task" value="load"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
255 <output name="outfile_predict" file="rfc_result01" compare="sim_size" delta="500"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
256 </test>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
257 <test>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
258 <param name="infile1" value="regression_train.tabular" ftype="tabular"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
259 <param name="infile2" value="regression_train.tabular" ftype="tabular"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
260 <param name="col1" value="1,2,3,4,5"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
261 <param name="col2" value="6"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
262 <param name="selected_task" value="train"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
263 <param name="selected_algorithm" value="RandomForestRegressor"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
264 <param name="random_state" value="10"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
265 <output name="outfile_fit" file="rfr_model01" compare="sim_size" delta="500"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
266 </test>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
267 <test>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
268 <param name="infile_model" value="rfr_model01" ftype="zip"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
269 <param name="infile_data" value="regression_test.tabular" ftype="tabular"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
270 <param name="selected_task" value="load"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
271 <output name="outfile_predict" file="rfr_result01" compare="sim_size" delta="500"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
272 </test>
5
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
273 <test>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
274 <param name="infile1" value="regression_X.tabular" ftype="tabular"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
275 <param name="infile2" value="regression_y.tabular" ftype="tabular"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
276 <param name="header1" value="True"/>
10
cd595710f0c0 planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit 97c4f22cdcfa6cddeeffc7b102c418a7ff12a888
bgruening
parents: 5
diff changeset
277 <param name="selected_column_selector_option" value="all_columns"/>
5
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
278 <param name="header2" value="True"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
279 <param name="col2" value="1"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
280 <param name="selected_task" value="train"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
281 <param name="selected_algorithm" value="GradientBoostingRegressor"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
282 <param name="max_features" value="number_input"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
283 <param name="num_max_features" value=""/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
284 <param name="random_state" value="42"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
285 <output name="outfile_fit" file="gbr_model01" compare="sim_size" delta="500"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
286 </test>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
287 <test>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
288 <param name="infile_model" value="gbr_model01" ftype="zip"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
289 <param name="infile_data" value="regression_test_X.tabular" ftype="tabular"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
290 <param name="selected_task" value="load"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
291 <param name="header" value="True"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
292 <output name="outfile_predict" file="gbr_prediction_result01.tabular" compare="sim_size" delta="500"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
293 </test>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
294 <test>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
295 <param name="infile1" value="train.tabular" ftype="tabular"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
296 <param name="infile2" value="train.tabular" ftype="tabular"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
297 <param name="col1" value="1,2,3,4"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
298 <param name="col2" value="5"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
299 <param name="selected_task" value="train"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
300 <param name="selected_algorithm" value="GradientBoostingClassifier"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
301 <output name="outfile_fit" file="gbc_model01" compare="sim_size" delta="500"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
302 </test>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
303 <test>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
304 <param name="infile_model" value="gbc_model01" ftype="zip"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
305 <param name="infile_data" value="test.tabular" ftype="tabular"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
306 <param name="selected_task" value="load"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
307 <output name="outfile_predict" file="gbc_result01" compare="sim_size" delta="500"/>
f1761288587e planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 35fa73d6e9ba8f0789ddfb743d893d950a68af02
bgruening
parents: 4
diff changeset
308 </test>
0
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
309 </tests>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
310 <help><![CDATA[
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
311 ***What it does***
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
312 The goal of ensemble methods is to combine the predictions of several base estimators built with a given learning algorithm in order to improve generalizability / robustness over a single estimator. This tool offers two sets of ensemble algorithms for classification and regression: random forests and ADA boosting which are based on sklearn.ensemble library from Scikit-learn. Here you can find out about the input, output and methods presented in the tools. For information about ensemble methods and parameters settings please refer to `Scikit-learn ensemble`_.
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
313
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
314 .. _`Scikit-learn ensemble`: http://scikit-learn.org/stable/modules/ensemble.html
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
315
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
316 **1 - Methods**
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
317 There are two groups of operations available:
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
318
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
319 1 - Train a model : A training set containing samples and their respective labels (or predicted values) are input. Based on the selected algorithm and options, an estimator object is fit to the data and is returned.
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
320
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
321 2 - Load a model and predict : An existing model predicts the class labels (or regression values) for a new dataset.
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
322
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
323 **2 - Trainig input**
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
324 When you choose to train a model, you need a features dataset X and a labels set y. This tool expects tabular or sparse data for X and a single column for y (tabular). You can select a subset of columns in a tabular dataset as your features dataset or labels column. Below you find some examples:
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
325
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
326 **Sample tabular features dataset**
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
327 The following training dataset contains 3 feature columns and a column containing class labels. You can simply select the first 3 columns as features and the last column as labels:
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
328
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
329 ::
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
330
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
331 4.01163365529 -6.10797684314 8.29829894763 1
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
332 10.0788438916 1.59539821454 10.0684278289 0
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
333 -5.17607775503 -0.878286135332 6.92941850665 2
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
334 4.00975406235 -7.11847496542 9.3802423585 1
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
335 4.61204065139 -5.71217537352 9.12509610964 1
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
336
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
337
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
338 **Sample sparse features dataset**
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
339 In this case you cannot specifiy a column range.
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
340
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
341 ::
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
342
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
343 4 1048577 8738
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
344 1 271 0.02083333333333341
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
345 1 1038 0.02461995616119806
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
346 2 829017 0.01629088031127686
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
347 2 829437 0.01209127083516686
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
348 2 830752 0.02535100632816968
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
349 3 1047487 0.01485722929945572
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
350 3 1047980 0.02640566620767753
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
351 3 1048475 0.01665869913262564
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
352 4 608 0.01662975263094352
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
353 4 1651 0.02519674277562741
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
354 4 4053 0.04223659971350601
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
355
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
356
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
357 **2 - Trainig output**
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
358 The trained model is generated and output in the form of a binary file.
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
359
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
360
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
361 **3 - Prediction input**
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
362
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
363 When you choose to load a model and do prediction, the tool expects an already trained estimator and a tabular dataset as input. The dataset contains new samples which you want to classify or predict regression values for.
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
364
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
365
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
366 .. class:: warningmark
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
367
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
368 The number of feature columns must be the same in training and prediction datasets!
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
369
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
370
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
371 **3 - Prediction output**
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
372 The tool predicts the class labels for new samples and adds them as the last column to the prediction dataset. The new dataset then is output as a tabular file. The prediction output format should look like the training dataset.
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
373
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
374 ]]></help>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
375 <expand macro="sklearn_citation"/>
569eefee7ed8 planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 0e582cf1f3134c777cce3aa57d71b80ed95e6ba9
bgruening
parents:
diff changeset
376 </tool>