Closed DecimalTurn closed 1 year ago
Exactly! And this new project (blue line) can be roughly 10x faster to read data compared to the other CSV project already listed (from ws-garcia).
Actually reading more into it, not so sure. Garcia raised an issue which was never looked at by the authors by the looks of it
You're right, I missed that issueπ . At least the project still has an interesting "approach of analyzing and importing CSV files from Excel formulas", so it might still have a good reason to have its place in the awesome-vba list.
Actually reading more into it, not so sure. Garcia raised an issue which was never looked at by the authors by the looks of it
After opening a second issue, explaining in more depth the TRUE PERFORMANCE of VBA CSV and putting into context the memory usage and increased runtimes when working with really large files, the author's response was to remove the issues and also the performance section from the Readme
file.
A "awesome" library, whose author does not admit the weaknesses/strengths of its creation, which also does not respond adequately to the issues launched in detail, only the inadequate attitude of its creator is awesome.
Yikes... Deleting the issue is pretty much the worst thing that could be done... Now there is no record of performance comparrison at all... π I honestly didn't know deleting issues was even possible... Awful VC...
Edit: I'm going to be reverting this addition. On the basis that it's:
Irrespective of the author's actions.
Yikes... Deleting the issue is pretty much the worst thing that could be done... Now there is no record of performance comparrison at all... π I honestly didn't know deleting issues was even possible... Awful VC...
Haha, did not realize it's possible. Like you, I never did that to anyone - it's just rude
Appears to have no performance benefits
As I have lost all the information I posted in the published issues, here is a table with the data on which I based the last issue in which I refuted the 14 MB/sec. processing speed that the author of VBA CSV alluded to in the Readme
file.
Tool | File | File size (MB) | Parsing time (seconds) | Speed (MB/sec.) |
---|---|---|---|---|
CSV Interface | 100000.quoted.csv | 11.8 | 3.9062 | 3.02 |
VBA CSV | 100000.quoted.csv | 11.8 | 4.4219 | 2.67 |
CSV Interface | 400000.quoted.csv | 47.5 | 11.8867 | 3.99 |
VBA CSV | 400000.quoted.csv | 47.5 | 13.7305 | 3.46 |
CSV Interface | 800000.quoted.csv | 95.1 | 22.9531 | 4.14 |
VBA CSV | 800000.quoted.csv | 95.1 | 27.1133 | 3.51 |
CSV Interface | 1600000.quoted.csv | 190 | 47.1953 | 4.02 |
VBA CSV | 1600000.quoted.csv | 190 | 60.793 | 3.12 |
CSV Interface | 3200000.quoted.csv | 383 | 98.6328 | 3.88 |
VBA CSV | 3200000.quoted.csv | 383 | 130.8008 | 2.92 |
CSV Interface | 6400000.quoted.csv | 767 | 209.8438 | 3.66 |
VBA CSV | 6400000.quoted.csv | 767 | Out of string memory error! | - |
Yikes... Deleting the issue is pretty much the worst thing that could be done... Now there is no record of performance comparrison at all... π I honestly didn't know deleting issues was even possible... Awful VC...
Edit: I'm going to be reverting this addition. On the basis that it's:
- Excel only
- Windows only
- Appears to have no performance benefits
Irrespective of the author's actions.
For the record, I'm OK with this decision.
Thanks for the PR, definitely worth the add especially as they have performance checks