Streams optional execution paths depending upon data rates

A method of processing data comprising:

• receiving streaming data to be processed by a plurality of interconnected processing elements, each processing element comprising one or more operators that process at least a portion of the received data by operation of one or more computer processors, wherein each one of the plurality of interconnected processing elements is hosted on a corresponding compute node;
• measuring, during a first time period, a data flow rate in a data path between at least two operators in the plurality of processing elements processing the streaming data;
• processing, during the first time period, at least a portion of the streaming data using a first code module, wherein the streaming data comprises a plurality of data tuples where each of the plurality of data tuples comprises a plurality of attribute value pairs, wherein the first code module processes a first attribute value pair of the plurality of attribute value pairs;
• selecting, based on the measured data flow rate, an inactive code module, wherein the selected code module is maintained in an inactive state until the data flow rate satisfies a predefined threshold, wherein the data flow rate is at least one of the number of data elements flowing in the data path during a predefined time period or a ratio of ingress data elements to egress data elements;
• upon determining that the data flow rate satisfies the predefined threshold, fusing an operator to a first one of the plurality of processing elements, wherein the fused operator comprises the inactive code module;
• activating, during a second time period, the selected code module on the first plurality of processing element such that a second attribute value pair of the plurality of attribute value pairs in the streaming data received by the first processing element is processed by the selected code module, wherein the second time period occurs after the first time period, wherein the first code module processes the first attribute value pair during the second time period, and wherein the first attribute value pair is different from the second attribute value pair;
• and
• upon determining that the data flow rate no longer satisfies the predefined threshold, un-fusing the fused operator from the first processing element, thereby removing the selected code module from the first processing element.