bigmlcom / bigml-php Goto Github PK

I believe there's a couple of errors in

https://github.com/bigmlcom/bigml-php/blob/master/bigml/ensemble.php#L111

that I think should be changed to

          foreach (range(0, $number_of_models - 1, $max_models) as $index) {
              array_push($this->models_splits,
                         array_slice($models, $index, $max_models));
          }

and also in multivote extend method's check (the argument can be a stdClass object).

https://github.com/bigmlcom/bigml-php/blob/master/bigml/multivote.php#L857

so it should be changed to something like:

   function extend($predictions_info) {
      /*Given a list of predictions, extends the list with another list of
           predictions and adds the order information. For instance,
           predictions_info could be:

                [{'prediction': 'Iris-virginica', 'confidence': 0.3},
                 {'prediction': 'Iris-versicolor', 'confidence': 0.8}]
           where the expected prediction keys are: prediction (compulsory),
           confidence, distribution and count.
       */
       if (is_array($predictions_info) ) {
         $order = $this->next_order();
         $i=0;
         foreach($predictions_info as $prediction) {
	    if (is_array($prediction)) {
	      $prediction['order'] = $order+$i;
	      array_push($this->predictions, $prediction);
	    } else if (get_class($prediction) == 'stdClass') {
          $prediction->order = $order + $i;
	      array_push($this->predictions, $prediction);
        } else {
	       error_log("WARNING: failed to add the prediction.\n Only dict like predictions are expected\n");
	    }
            $i+=1;
         }
       } else {
         error_log("WARNING: failed to add the predictions.\nOnly a list of dict-like predictions are expected.");
       }
   }

They prevent the ensemble predict method from working when the max_models is set to some integer lower than the number of models in the ensemble. For a 10 model ensemble you can test it like this.

php > include 'ensemble.php';
php > $ensemble = new Ensemble('ensemble/52faea8c035d07258d00861f', null, 2);
php > $ensemble->predict(array());

Some late changes in the Logistic Regression arguments need to be incorporated in the Local Logistic Regression object. Fix test_scenario12 and test_scenario14 in test_05_compare_predictions.

My Current PHP version: 7.1.28
My PHPUnit 7.5.9 by Sebastian Bergmann and contributors.
PHPUnit installed with PHP Archive (PHAR)
➜ wget -O phpunit https://phar.phpunit.de/phpunit-7.phar
➜ chmod +x phpunit
➜ ./phpunit --version

but after installing PHPUnit with PHAR it isn't generate "src/autoload.php" but just only file phpunit

but if I run ./phpunit --version in terminal it's show PHPUnit 7.5.9 by Sebastian Bergmann and contributors.

I download the zip in https://github.com/bigmlcom/bigml-php and extract it then I run it bigml-php-master/tests/test_00_regressions.php then I got an error message PHP Fatal error: Class 'PHPUnit_Framework_TestCase' not found

can anybody help me please

how i can change the default objective field?
thanks

As per a user request, we should add the equivalent of the api.ok method in the Python bindings

Existe un error en la línea 115 del fichero bigml.php.

El parámetro que recibe la función no es el que se asigna al static.

Debería ser $devMode y no $defMode.

public static function setDevMode($devMode=false) {
self::$devMode = $defMode;
}

Un saludo y gracias.