Provide additional resources
Zholdosh
For several days in a row I have been trying to classify. I have a territory about 1.5 million hectares. Created standard 6 classes of the land cover. For the training of the model, I have over 1100 points. But in spite of all this, for some reason, it doesn’t work out. Could it be related to limited resources? If so, I ask you to provide additional resources.
Teo Nakalema
Zholdosh
Dear Teo,
The issue has been resolved. It was related to the CSV file containing training data for approximately 1100 points. No errors were detected within the SEPAL interface itself; the error was identified in the Google Earth Engine interface:
Error: A CSV row had more columns than the header in file 'TAreaCEarthGri_FeatureToPoin_ExportTable4TrainData.csv': row size 3 > header size 1 for a row starting with: "74", "599707;41", "44;1" (Error code: 3)
The problem was due to the use of incorrect delimiters. Instead of semicolons, I had used commas. After correcting this and rerunning the classification, everything worked fine:
Xcoordinate;YCoordinate;class 74.599707;41.44;1
I would also like to highlight another point. In previous classifications, when I manually marked the training data, I faced similar issues. I suspected that it might be due to insufficient computational resources for such a large area. After requesting and obtaining more resources, the classification of an area of about 1.6 million hectares was successful. At the time of classification, I had the following resources: Instance - $50.0/mo, Storage spending - $50.0/mo.
Therefore, the two key factors in resolving the issue were correcting the delimiters in the CSV file and increasing the computational resources.
Thank you for your response!