csvgraph 2v9

2v9 – 7/6/2022 – bug fix – if x-offset is not equal to 0 and multiple traces added incorrect x-offset
was applied to the 2nd trace added onwards.
Higher resolution internally on reading dates & times, so if “start time from zero” is
ticked the results may be slightly more accurate.

csvgraph 2v8

Added ability to read a date and time as the x value.

csvsave added % complete and (significantly) sped up writing to file.

csvsave interpolates if required so x values do not need to be identical on all traces.

csvgraph 2v7

Prints -3dB frequency for linear filter.
Display to user at least 1 example of every type of error in csv file (used to only display the first few errors found).
If dates are present on some lines of the csv file then flag lines without a date as a potential error when x column is set to date/time (these lines are still loaded in the same way as they were previoulsy).
New (exact) median (recursive median filter) algorithm, which falls back to sampling if the execution time becomes long. Median and Median 1 renamed as recursive median filter and standard median filter (see help/manual for more details on these median filters).

csvgraph 2v6

2v6 – 27/2/2022 – Median1 filter improved, for <=10,000 data points its now exact and it's (much) more for accurate traces with a larger number of data points.
Position of trace legends moved left so more text can be seen.

csvgraph 2v5

Better reporting of errors in csv files.
When times read, can optionally display graph starting at time=0
Minor other improvements/bug fixes, especially when reading files > 2GB where previoulsy % complete values could be inaccurate.

csvgraph 2v4

Csvgraph 2v4 is released to fix a bug: using variables $Tn in an expression with a set of x-axis values that were
not in numerically increasing order (and so needed to be sorted) did not work correctly in 2v2 and 2v3 – sorry.

csvgraph 2v3

Version 2v3 of csvgraph released on 23/1/2022
Added y=mxlog(x)+c curve fitting. Sort and median functions improved to make them faster - sort will use all available processors to improve its execution speed.

csvgraph 2v2

csvgraph 2v2 - 5/4/2021

$T1 to Tn allowed in expressions to use values from existing traces on the graph.
Traces are numbered from 1. Invalid trace numbers (too big) return 0.

User can now set order of the linear filter. This is implemented as a nth order Butterworth filter (10*order dB/decade). Order=0 gives no filtering. Order =1 gives same filtering as previously.

"filters" for integral and derivative added.

All filters now report progress as a % (previously min. abs error and min. relative error did not report progress and they can be quite slow).

Option to skip N lines before csv header added for cases where csv header is not on the 1st row of the file.

Added column numbers to X column and Y column list boxes to make it easier to select columns when names are not very descriptive (or missing).

Csvgraph 2v1

2v1 – 21/3/2021 – more curve fitting options added; y=a+bsqrt(x)+cx+d*x^1.5, y=(a+bx)/(1+cx),
y=(a+bx+cx^2)/(1+dx+ex^2), polynomials on sqrt(x) of user defined order and rational functions (polynomial/polynomial) to a user defined order.

csvgraph 2v0

Major internal changes to reduce RAM usage and improve speed.

No changes to function.