« blib/lib/Paws/MachineLearning/RedshiftMetadata.pm
↑ Summary
blib/lib/Paws/MachineLearning/ScoreValuePerLabelMap.pm »
| 1 | # Generated by default/object.tt | ||
| 2 | package Paws::MachineLearning::S3DataSpec; | ||
| 3 | use Moose; | ||
| 4 | has DataLocationS3 => (is => 'ro', isa => 'Str', required => 1); | ||
| 5 | has DataRearrangement => (is => 'ro', isa => 'Str'); | ||
| 6 | has DataSchema => (is => 'ro', isa => 'Str'); | ||
| 7 | has DataSchemaLocationS3 => (is => 'ro', isa => 'Str'); | ||
| 8 | |||
| 9 | 1; | ||
| 10 | |||
| 11 | ### main pod documentation begin ### | ||
| 12 | |||
| 13 | =head1 NAME | ||
| 14 | |||
| 15 | Paws::MachineLearning::S3DataSpec | ||
| 16 | |||
| 17 | =head1 USAGE | ||
| 18 | |||
| 19 | This class represents one of two things: | ||
| 20 | |||
| 21 | =head3 Arguments in a call to a service | ||
| 22 | |||
| 23 | Use the attributes of this class as arguments to methods. You shouldn't make instances of this class. | ||
| 24 | Each attribute should be used as a named argument in the calls that expect this type of object. | ||
| 25 | |||
| 26 | As an example, if Att1 is expected to be a Paws::MachineLearning::S3DataSpec object: | ||
| 27 | |||
| 28 | $service_obj->Method(Att1 => { DataLocationS3 => $value, ..., DataSchemaLocationS3 => $value }); | ||
| 29 | |||
| 30 | =head3 Results returned from an API call | ||
| 31 | |||
| 32 | Use accessors for each attribute. If Att1 is expected to be an Paws::MachineLearning::S3DataSpec object: | ||
| 33 | |||
| 34 | $result = $service_obj->Method(...); | ||
| 35 | $result->Att1->DataLocationS3 | ||
| 36 | |||
| 37 | =head1 DESCRIPTION | ||
| 38 | |||
| 39 | Describes the data specification of a C<DataSource>. | ||
| 40 | |||
| 41 | =head1 ATTRIBUTES | ||
| 42 | |||
| 43 | |||
| 44 | =head2 B<REQUIRED> DataLocationS3 => Str | ||
| 45 | |||
| 46 | The location of the data file(s) used by a C<DataSource>. The URI | ||
| 47 | specifies a data file or an Amazon Simple Storage Service (Amazon S3) | ||
| 48 | directory or bucket containing data files. | ||
| 49 | |||
| 50 | |||
| 51 | =head2 DataRearrangement => Str | ||
| 52 | |||
| 53 | A JSON string that represents the splitting and rearrangement | ||
| 54 | processing to be applied to a C<DataSource>. If the | ||
| 55 | C<DataRearrangement> parameter is not provided, all of the input data | ||
| 56 | is used to create the C<Datasource>. | ||
| 57 | |||
| 58 | There are multiple parameters that control what data is used to create | ||
| 59 | a datasource: | ||
| 60 | |||
| 61 | =over | ||
| 62 | |||
| 63 | =item * | ||
| 64 | |||
| 65 | B<C<percentBegin>> | ||
| 66 | |||
| 67 | Use C<percentBegin> to indicate the beginning of the range of the data | ||
| 68 | used to create the Datasource. If you do not include C<percentBegin> | ||
| 69 | and C<percentEnd>, Amazon ML includes all of the data when creating the | ||
| 70 | datasource. | ||
| 71 | |||
| 72 | =item * | ||
| 73 | |||
| 74 | B<C<percentEnd>> | ||
| 75 | |||
| 76 | Use C<percentEnd> to indicate the end of the range of the data used to | ||
| 77 | create the Datasource. If you do not include C<percentBegin> and | ||
| 78 | C<percentEnd>, Amazon ML includes all of the data when creating the | ||
| 79 | datasource. | ||
| 80 | |||
| 81 | =item * | ||
| 82 | |||
| 83 | B<C<complement>> | ||
| 84 | |||
| 85 | The C<complement> parameter instructs Amazon ML to use the data that is | ||
| 86 | not included in the range of C<percentBegin> to C<percentEnd> to create | ||
| 87 | a datasource. The C<complement> parameter is useful if you need to | ||
| 88 | create complementary datasources for training and evaluation. To create | ||
| 89 | a complementary datasource, use the same values for C<percentBegin> and | ||
| 90 | C<percentEnd>, along with the C<complement> parameter. | ||
| 91 | |||
| 92 | For example, the following two datasources do not share any data, and | ||
| 93 | can be used to train and evaluate a model. The first datasource has 25 | ||
| 94 | percent of the data, and the second one has 75 percent of the data. | ||
| 95 | |||
| 96 | Datasource for evaluation: C<{"splitting":{"percentBegin":0, | ||
| 97 | "percentEnd":25}}> | ||
| 98 | |||
| 99 | Datasource for training: C<{"splitting":{"percentBegin":0, | ||
| 100 | "percentEnd":25, "complement":"true"}}> | ||
| 101 | |||
| 102 | =item * | ||
| 103 | |||
| 104 | B<C<strategy>> | ||
| 105 | |||
| 106 | To change how Amazon ML splits the data for a datasource, use the | ||
| 107 | C<strategy> parameter. | ||
| 108 | |||
| 109 | The default value for the C<strategy> parameter is C<sequential>, | ||
| 110 | meaning that Amazon ML takes all of the data records between the | ||
| 111 | C<percentBegin> and C<percentEnd> parameters for the datasource, in the | ||
| 112 | order that the records appear in the input data. | ||
| 113 | |||
| 114 | The following two C<DataRearrangement> lines are examples of | ||
| 115 | sequentially ordered training and evaluation datasources: | ||
| 116 | |||
| 117 | Datasource for evaluation: C<{"splitting":{"percentBegin":70, | ||
| 118 | "percentEnd":100, "strategy":"sequential"}}> | ||
| 119 | |||
| 120 | Datasource for training: C<{"splitting":{"percentBegin":70, | ||
| 121 | "percentEnd":100, "strategy":"sequential", "complement":"true"}}> | ||
| 122 | |||
| 123 | To randomly split the input data into the proportions indicated by the | ||
| 124 | percentBegin and percentEnd parameters, set the C<strategy> parameter | ||
| 125 | to C<random> and provide a string that is used as the seed value for | ||
| 126 | the random data splitting (for example, you can use the S3 path to your | ||
| 127 | data as the random seed string). If you choose the random split | ||
| 128 | strategy, Amazon ML assigns each row of data a pseudo-random number | ||
| 129 | between 0 and 100, and then selects the rows that have an assigned | ||
| 130 | number between C<percentBegin> and C<percentEnd>. Pseudo-random numbers | ||
| 131 | are assigned using both the input seed string value and the byte offset | ||
| 132 | as a seed, so changing the data results in a different split. Any | ||
| 133 | existing ordering is preserved. The random splitting strategy ensures | ||
| 134 | that variables in the training and evaluation data are distributed | ||
| 135 | similarly. It is useful in the cases where the input data may have an | ||
| 136 | implicit sort order, which would otherwise result in training and | ||
| 137 | evaluation datasources containing non-similar data records. | ||
| 138 | |||
| 139 | The following two C<DataRearrangement> lines are examples of | ||
| 140 | non-sequentially ordered training and evaluation datasources: | ||
| 141 | |||
| 142 | Datasource for evaluation: C<{"splitting":{"percentBegin":70, | ||
| 143 | "percentEnd":100, "strategy":"random", | ||
| 144 | "randomSeed"="s3://my_s3_path/bucket/file.csv"}}> | ||
| 145 | |||
| 146 | Datasource for training: C<{"splitting":{"percentBegin":70, | ||
| 147 | "percentEnd":100, "strategy":"random", | ||
| 148 | "randomSeed"="s3://my_s3_path/bucket/file.csv", "complement":"true"}}> | ||
| 149 | |||
| 150 | =back | ||
| 151 | |||
| 152 | |||
| 153 | |||
| 154 | =head2 DataSchema => Str | ||
| 155 | |||
| 156 | A JSON string that represents the schema for an Amazon S3 | ||
| 157 | C<DataSource>. The C<DataSchema> defines the structure of the | ||
| 158 | observation data in the data file(s) referenced in the C<DataSource>. | ||
| 159 | |||
| 160 | You must provide either the C<DataSchema> or the | ||
| 161 | C<DataSchemaLocationS3>. | ||
| 162 | |||
| 163 | Define your C<DataSchema> as a series of key-value pairs. C<attributes> | ||
| 164 | and C<excludedVariableNames> have an array of key-value pairs for their | ||
| 165 | value. Use the following format to define your C<DataSchema>. | ||
| 166 | |||
| 167 | { "version": "1.0", | ||
| 168 | |||
| 169 | "recordAnnotationFieldName": "F1", | ||
| 170 | |||
| 171 | "recordWeightFieldName": "F2", | ||
| 172 | |||
| 173 | "targetFieldName": "F3", | ||
| 174 | |||
| 175 | "dataFormat": "CSV", | ||
| 176 | |||
| 177 | "dataFileContainsHeader": true, | ||
| 178 | |||
| 179 | "attributes": [ | ||
| 180 | |||
| 181 | { "fieldName": "F1", "fieldType": "TEXT" }, { "fieldName": "F2", | ||
| 182 | "fieldType": "NUMERIC" }, { "fieldName": "F3", "fieldType": | ||
| 183 | "CATEGORICAL" }, { "fieldName": "F4", "fieldType": "NUMERIC" }, { | ||
| 184 | "fieldName": "F5", "fieldType": "CATEGORICAL" }, { "fieldName": "F6", | ||
| 185 | "fieldType": "TEXT" }, { "fieldName": "F7", "fieldType": | ||
| 186 | "WEIGHTED_INT_SEQUENCE" }, { "fieldName": "F8", "fieldType": | ||
| 187 | "WEIGHTED_STRING_SEQUENCE" } ], | ||
| 188 | |||
| 189 | "excludedVariableNames": [ "F6" ] } | ||
| 190 | |||
| 191 | |||
| 192 | =head2 DataSchemaLocationS3 => Str | ||
| 193 | |||
| 194 | Describes the schema location in Amazon S3. You must provide either the | ||
| 195 | C<DataSchema> or the C<DataSchemaLocationS3>. | ||
| 196 | |||
| 197 | |||
| 198 | |||
| 199 | =head1 SEE ALSO | ||
| 200 | |||
| 201 | This class forms part of L<Paws>, describing an object used in L<Paws::MachineLearning> | ||
| 202 | |||
| 203 | =head1 BUGS and CONTRIBUTIONS | ||
| 204 | |||
| 205 | The source code is located here: L<https://github.com/pplu/aws-sdk-perl> | ||
| 206 | |||
| 207 | Please report bugs to: L<https://github.com/pplu/aws-sdk-perl/issues> | ||
| 208 | |||
| 209 | =cut | ||
| 210 |