Hello,
Has anyone gotten the tar.gz files to open or work in Alteryx using the input tool? I believe the files inside of it are csv and that may be the reason that it is not working? Any suggestions or experience?
I have tried this article and it did not help: https://community.alteryx.com/t5/Alteryx-Designer-Discussions/Extracting-files-from-zip-files-or-tar...
Thanks in advance!
Hi @GIS_User01,
I wouldn't call myself an expert in this area, but I just tried it with a csv file inside and it worked right away. Could you maybe try to use something like 7zip to check what's inside?
Btw. it looks a bit different when you drag and drop it into the canvas:
But after that it looks rather similar in the Input tool and you have all the options you need:
What does it look on your end?
I might be wrong, but I think I saw that you also worked with these type of files, weren't you @MarqueeCrew ?
Best
Alex
@grossal ,
what you've shown here looks right. Two thoughts: do you have a zip extractor loaded on your machine (7zip) and you might want UTF8.
then the simple check for the right delimiter and default field length.
cheers,
mark
ps if the files are large, AMP isn't a good fit.
Thanks to you both - @grossal & @MarqueeCrew
It turns out that the file I was sent had some sort of corrupt row that kind of messed up the whole dataset. There are around 124 million records but Alteryx only shows the first 32 million or so due to that error.
I will be receiving a re-delivery of the data and I am hoping the process will run smoothly with the new dataset!
Thanks!
Hi @GIS_User01,
that error does not necessarily mean that something is wrong with the CSV in my experience. I would recommend the following:
Import the CSV with the Delimiter \0 (no Delimiter).
Afterwards you'll see that every row is just one cell instead of multiple. Therefore we can use a Text-to-Columns tool (comma as delimiter) and spread the columns out normally. This usually resolves the issue.
Best
Alex
Thanks again @grossal (Alex)!
I gave that a shot (changing the delimiter to "\0" and got some errors again:
I am expecting a re-delivery of the data today.
Thanks again,
Bobby
Ahh, I am sorry! I forgot to tell that we also need to increase the 'Field length'. I usually use something like 10.000, maybe 100k if I know that the text is long, but than I see the last error. It seems you were right, there is an issue in the data. A field that is larger than 16M - 16 Million characters seems odd.
User | Count |
---|---|
19 | |
15 | |
13 | |
9 | |
8 |