Methods, systems, and computer program products for a multi-resolution storage scheme for historical data

Abstract

A method, system, and computer program product for a multi-resolution storage scheme for historical data are provided. The method includes dividing the historical data into a plurality of time frames and determining an amount of compression for each of the time frames, where at least two of the time frames have a different amount of compression. The method further includes applying the amount of compression to each of the time frames, resulting in multi-resolution historical data, and storing the multi-resolution historical data.

1 . A method for a multi-resolution storage scheme for historical data, comprising:
 dividing historical data into a plurality of time frames;   determining an amount of compression for each of the time frames, wherein at least two of the time frames have a different amount of compression;   applying the amount of compression to each of the time frames, resulting in multi-resolution historical data; and   storing the multi-resolution historical data.   
   
   
       2 . The method of  claim 1  wherein a greater amount of compression is applied to older time frames than to younger time frames. 
   
   
       3 . The method of  claim 1  further comprising:
 compressing a plurality of series of the historical data in multiple dimensions, wherein older series of the historical data are compressed as a two-dimensional image, and younger series of the historical data are compressed in a single dimension.   
   
   
       4 . The method of  claim 1  wherein the compression is applied via a wavelet transform and output as wavelet coefficients. 
   
   
       5 . The method of  claim 4  wherein the wavelet coefficients are generated as one or more levels of approximation and detail coefficients. 
   
   
       6 . The method of  claim 4  wherein the wavelet coefficients are compressed via a lossless encoding technique. 
   
   
       7 . The method of  claim 4  further comprising:
 merging a plurality of compressed time frames via a time series merger with partial reconstruction.   
   
   
       8 . The method of  claim 4  further comprising:
 removing historical data from the multi-resolution historical data via identifying the wavelet coefficients associated with a portion of the historical data to retain, and discarding remaining unidentified wavelet coefficients.   
   
   
       9 . The method of  claim 4  further comprising:
 reconstructing the historical data through applying an inverse wavelet transform to the wavelet coefficients.   
   
   
       10 . The method of  claim 9  further comprising:
 calculating an error metric as a mean square error between the historical data and the reconstructed historical data; and   outputting the error metric.   
   
   
       11 . The method of  claim 1  wherein the amount of compression is calculated based on one of a requested error tolerance and a space constraint for storing the multi-resolution historical data. 
   
   
       12 . A system for a multi-resolution storage scheme for historical data, comprising:
 a database; and   a host system in communication with the database, the host system executing a compression layer (CL), the CL performing:
 receiving historical data; 
 dividing the historical data into a plurality of time frames; 
 determining an amount of compression for each of the time frames, wherein at least two of the time frames have a different amount of compression; 
 applying the amount of compression to each of the time frames, resulting in multi-resolution historical data; and 
 storing the multi-resolution historical data in the database. 
   
   
   
       13 . The system of  claim 12  wherein a greater amount of compression is applied to older time frames than to younger time frames. 
   
   
       14 . The system of  claim 12  wherein the compression is applied via a wavelet transform and output as wavelet coefficients. 
   
   
       15 . The system of  claim 14  wherein the wavelet coefficients are compressed via a lossless encoding technique. 
   
   
       16 . The system of  claim 14  wherein the CL further performs:
 merging compressed time frames via a time series merger with partial reconstruction; and   removing historical data from the multi-resolution historical data via identifying the wavelet coefficients associated with a portion of the historical data to retain, and discarding remaining unidentified wavelet coefficients.   
   
   
       17 . The system of  claim 14  wherein the CL further performs:
 reconstructing the historical data through applying an inverse wavelet transform to the wavelet coefficients.   
   
   
       18 . The system of  claim 17  wherein the CL further performs:
 calculating an error metric as a mean square error between the historical data and the reconstructed historical data; and   outputting the error metric.   
   
   
       19 . A computer program product for a multi-resolution storage scheme for historical data, the computer program product comprising:
 a storage medium readable by a processing circuit and storing instructions for execution by the processing circuit for implementing a method, the method comprising:
 dividing historical data into a plurality of time frames; 
 determining an amount of compression for each of the time frames, wherein at least two of the time frames have a different amount of compression; 
 applying the amount of compression to each of the time frames, resulting in multi-resolution historical data; and 
 storing the multi-resolution historical data. 
   
   
   
       20 . The computer program product of  claim 19  wherein a greater amount of compression is applied to older time frames than to younger time frames.