Upload file force same filename

Hello,

A file is renamed in an random way. Correct?

Example:

If I upload test_2.9.csv using the dashboard I obtain test_2.9.csv.2hpla6of

This is something I like because I can link filename test_2.9.csv.2hpla6of with the file test_2.9.csv on my local machine. (note 2.9 is my speed label in m/s)

However using edge-impulse-uploader I obtain
test_2.2hpm7vod
in this case the .9.csv is thrown away (however this is not in case I use the dashboard, as explained above).

In case of test_2p9.csv I obtain test_2p9.2hpmmm37, only csv is throw away.

I can use xpy instead of x.y as in between solution and write a Python script where i replace p with . for the label.

Currently I use

edge-impulse-uploader --category {category} --label {label} --api-key {api_key} {path+file}

Is there an option for the edge-impulse-uploader to force the same filename test_2.9.csv becomes test_2.9.csv.+something random.

Regards,
Joeri

In meantime I play with the EI API.
Using JSON gives:

โ€˜x-file-nameโ€™: โ€˜test_2.9.csvโ€™
Uploaded file to Edge Impulse 200 bโ€™test_2.9.csv.2hre2o59.jsonโ€™

I go for this approach.

Question: Is there a specific reason why there is some random value (in this case 2hre2o59) add at the end of the filename.

Hello @Joeri,

You can have a look at this thread: Getting floating points labels from file name
It is not completely answering your questions but it is somehow similar :wink:

A file is renamed in an random way. Correct?

Correct

test_2 .2hpm7vod
in this case the .9.csv is thrown away (however this is not in case I use the dashboard, as explained above).

Yes, we take as the label what is before the first period, the rest we give it a random value to create unicity.

Regards,

Louis