IndexError: During execution of task S2L1CWCSInput: index -1 is out of bounds for axis 0 with size 0

Hi ,

I am getting this error while downloading my area image. When using a shape file of my own region, generated via “SI_LULC_pipeline.ipynb” notebook…
also i found similar issue /here …Anyone knows how to handle this problem…Many Thanks in advance!

Hi Niraj,

Can you share the coordinates of the area you are analysing? Seems to be an issue with an invalid bounding box.

Hello @devis.peressutti
Thanks for this…how can I handle bounding box issue?
I’m doing it for pune,India
Cordinates for pune 18.5204° N, 73.8567° E

Is there anything if I am doing wrong please guide me.

Hello,

The following code works:

from eolearn.io import S2L1CWCSInput
from sentinelhub import BBox, CRS

s2_l1c = S2L1CWCSInput('TRUE-COLOR-S2-L1C', resx='10m', resy='10m', maxcc=.5)
bbox = BBox(((73.82461, 18.54928),(73.90778, 18.48791)), crs=CRS.WGS84)

eop_s2_l1c = s2_l1c.execute(bbox=bbox, time_interval=('2019-01-05','2019-02-05'))

Coordinate of bounding box must be ((lng_top_left, lat_top_left), (lng_bottom_right, lat_bottom_right)), as written in the sentinelhub documentation.

1 Like

Hi
I’m looking for the changes in this Notebook file…can you please suggest me what changes i should make so i can get pune images from Notebook file…and if i did changes in the notebook file I’m getting error… Thanks

Hi
@devis.peressutti needs help

Hi @Niraj

To run the notebook for a custom area you would need an input polygon defining your region-of-interest (e.g. Pune), and labels to be used to train the machine learning algorithm. The input bounding box can be hard-coded to the coordinates of Pune.

Do you have labelling data available for Pune? If yes, which format is it?

For pune i tried to find the input polygonal…do you have any idea where i can find the polygonals for pune…or do you have available for another city in india…need your help… Thanks

eo-learn assumes you have the labelling data already available in some way.

You could try to get some land cover labels from CCI-2015 (https://maps.elie.ucl.ac.be/CCI/viewer/) or GLC30 (http://www.globallandcover.com/GLC30Download/index.aspx).

Hello @devis.peressutti
Actually after getting your reply started searching on it…i found this link
helpful and downloaded image from herebut image size was heavy so it took time…so i am requesting you for help in the next part where i am stuck…also my subscription is over as i’m student do you have any idea for renewing subscription or students plan…i am very thankful to you!

I am interested in ml pipeline related work…may be i contribute something from my side for such nice project

Hello @devis.peressutti
Actually after getting your views on it, i started searching on it…i found this link
helpful and downloaded image from [here] (ESA/CCI viewer)but image size was heavy so it took time…and all bands are not covered you can check

…so i am requesting you for help in the next part where i am stuck…not able to find right image…if i am doing something wrong please help me…i am very thankful to you!