Content-aware compression of data with reduced number of class codes to be encoded

1. A method for compressing at least one floating point number, said method comprising:

obtaining a plurality of said floating point numbers, wherein a given floating point number is represented using one or more bits to indicate a sign of said given floating point number and one or more additional bits to indicate an exponent at a given base and a significand of said given floating point number, wherein said significand has a length equal to a number of bits between a most significant bit of said significand and a least significant bit of said significand having a predefined binary value;

transforming, using at least one processing device, a plurality of exponents of said plurality of floating point numbers to generate a transformed exponent to reduce a number of distinct exponents in said plurality of floating point numbers to be encoded; and

encoding, using said at least one processing device, said given floating point number by encoding said transformed exponent and said length of said given floating point number as a single class code.

2. The method of claim 1 , wherein said transformation comprises one or more of a delta encoding and a modulo operation.

3. The method of claim 1 , further comprising the step of applying a linear prediction algorithm to said at least one floating point number to generate a prediction, wherein said linear prediction algorithm comprises one or more of a first derivative and a second derivative.

4. The method of claim 3 , wherein one of said first derivative and said second derivative are selected based on an analysis of a statistically significant portion of data to be compressed.

5. The method of claim 1 , further comprising the step of automatically selecting a number of output class codes that need to be encoded.

6. The method of claim 5 , further comprising the steps of determining an entropy of a current set of classes and determining whether said entropy has increased with respect to an entropy value computed in a previous iteration.

7. The method of claim 1 , wherein said encoding step generates a class code, sign and significand bits for said given floating point number, and wherein said method further comprises the step of applying an entropy encoder to only said class code.

8. The method of claim 1 , further comprising the step of decompressing said compressed given floating point number by decompressing the class codes encoded with said entropy encoder, applying a reverse transformation of said transformation to reduce said number of distinct exponents and restoring discarded bits based on said encoded length.

9. The method of claim 1 , wherein said given floating point number comprises one or more of seismic data, signal data and sensor data.

10. A computer program product for compressing at least one floating point number, comprising a non-transitory machine-readable storage medium having encoded therein executable code of one or more software programs, wherein the one or more software programs when executed by at least one processing device perform the following steps:

obtaining a plurality of said floating point numbers, wherein a given floating point number is represented using one or more bits to indicate a sign of said given floating point number and one or more additional bits to indicate an exponent at a given base and a significand of said given floating point number, wherein said significand has a length equal to a number of bits between a most significant bit of said significand and a least significant bit of said significand having a predefined binary value;

transforming, using said at least one processing device, a plurality of exponents of said plurality of floating point numbers to generate a transformed exponent to reduce a number of distinct exponents in said plurality of floating point numbers to be encoded; and

encoding, using said at least one processing device, said given floating point number by encoding said transformed exponent and said length of said given floating point number as a single class code.

11. The computer program product of claim 10 , further comprising the step of applying a linear prediction algorithm to said at least one floating point number to generate a prediction, wherein said linear prediction algorithm comprises one or more of a first derivative and a second derivative, and wherein one of said first derivative and said second derivative are selected based on an analysis of a statistically significant portion of data to be compressed.

12. The computer program product of claim 10 , further comprising the step of automatically selecting a number of output class codes that need to be encoded.

13. The computer program product of claim 10 , wherein said encoding step generates a class code, sign and significand bits for said given floating point number, and wherein said computer program product further comprises the step of applying an entropy encoder to only said class code.

14. The computer program product of claim 10 , further comprising the step of decompressing said compressed given floating point number by decompressing the class codes encoded with said entropy encoder, applying a reverse transformation of said transformation to reduce said number of distinct exponents and restoring discarded bits based on said encoded length.

15. A system for compressing at least one floating point number, comprising:

a memory; and

at least one processing device, coupled to the memory, operative to implement the following steps:

obtaining a plurality of said floating point numbers, wherein a given floating point number is represented using one or more bits to indicate a sign of said given floating point number and one or more additional bits to indicate an exponent at a given base and a significand of said given floating point number, wherein said significand has a length equal to a number of bits between a most significant bit of said significand and a least significant bit of said significand having a predefined binary value;

transforming, using said at least one processing device, a plurality of exponents of said plurality of floating point numbers using a transformation that provides to generate a transformed exponent to reduce a number of distinct exponents in said plurality of floating point numbers to be encoded; and

encoding, using said at least one processing device, said given floating point number by encoding said transformed exponent and said length of said given floating point number as a single class code.

16. The system of claim 15 , further comprising the step of applying a linear prediction algorithm to said at least one floating point number to generate a prediction, wherein said linear prediction algorithm comprises one or more of a first derivative and a second derivative, and wherein one of said first derivative and said second derivative are selected based on an analysis of a statistically significant portion of data to be compressed.

17. The system of claim 15 , further comprising the step of automatically selecting a number of output class codes that need to be encoded.

18. The system of claim 15 , wherein said encoding step generates a class code, sign and significand bits for said given floating point number, and wherein said system further comprises the step of applying an entropy encoder to only said class code.

19. The system of claim 15 , further comprising the step of decompressing said compressed given floating point number by decompressing the class codes encoded with said entropy encoder, applying a reverse transformation of said transformation to reduce said number of distinct exponents and restoring discarded bits based on said encoded length.

20. The system of claim 15 , wherein said given floating point number comprises one or more of seismic data, signal data and sensor data.