point8 / fixitfelix Goto Github PK

In the writing process each chunk is written to the corrected file as a single segment. This increaes the read out time heavily. The chunks could be aggregated first and then written as one big segment.

Each chunk of the data between copied areas is written to the output file individually. As a result, for small input parameters and large files, there is a lot of fragmentation which can reduce the file readout time in some orders of magnitude.

Therefore we want to be able to write several chunks at once if they are small (any criterium possible, e.g. a prefered chunk size that is neared in the writing process).

Another benefit could be a faster writing process of the corrected file because of less iterations in the final for loop.

To check the chosen parameters the complete TDMS file is imported which can take some time.
If the check fails, this process has to be repeated.

To solve this issue and improve the overall performance simultaneously, one could use a more specific read method, which will only read the data of "used" channels.

Often we have several measurements of the same type (e.g. same paramters). It would be great to provide a list of files or a folder to correct more than one file with our tool.