Doubt in Statisticsop under GPT tool

Please enter here to ask a question about any NASA Science related topics!
Post Reply
ivgbhavani
Posts: 16
Joined: Mon May 21, 2018 11:23 am America/New_York

Doubt in Statisticsop under GPT tool

by ivgbhavani » Tue Jan 28, 2020 6:07 am America/New_York

Hi every one,
I am trying to use GPT tool to generate statistics. I could generate statistics like below for entire input file.
# Region  Band  average  max_error  maximum  median  minimum  p90_threshold  p95_threshold  sigma  total
world  chl_oc2  0.4037  0.0852  85.1800  0.2555  0.0000  0.4259  0.5963  0.7945  138158
I have two questions
1. Is it possible to specify region of interest (ROI) using east, west,north, south
2. How to generate only mean for the ROI / entire input.

Thanks in advance
Bhavani

Tags:

OB SeaDAS - aynur
Subject Matter Expert
Subject Matter Expert
Posts: 32
Joined: Tue Dec 20, 2011 12:28 pm America/New_York

Doubt in Statisticsop under GPT tool

by OB SeaDAS - aynur » Tue Jan 28, 2020 6:22 pm America/New_York

Hi Bhavani,

You can use a shapefile to specify the ROI in the StatisticsOp operator if you have one. Otherwise, you need to get the ROI using the Subset operator first before running StatisticsOp. You can combine these two operators in an xml file for batch processing.

Can you elaborate on your second question? Do you mean to generate mean only for all bands or only for one band?

Aynur

ivgbhavani
Posts: 16
Joined: Mon May 21, 2018 11:23 am America/New_York

Doubt in Statisticsop under GPT tool

by ivgbhavani » Wed Jan 29, 2020 3:10 am America/New_York

Thanks Aynur
I could get statistics using shape file.
My second question is  Is it possible to find specifically mean for one band ?. I don't want to print other parameters like region, band, max_error, maximum etc.
And one more thing I want to know.
this gpt StatisticsOP will work only for few file formats like netcdf files ? Can I use it for hdf4 files ?

Thanks for your time.

OB SeaDAS - aynur
Subject Matter Expert
Subject Matter Expert
Posts: 32
Joined: Tue Dec 20, 2011 12:28 pm America/New_York

Doubt in Statisticsop under GPT tool

by OB SeaDAS - aynur » Wed Jan 29, 2020 4:03 pm America/New_York

You're welcome, Bhavani.

The parameters you get from the StatisticsOp are hardcoded, so you can't be selective. You can write a script to extract the desired parameters from the output for the time being. We can consider adding that capacity to this operator in the future.

You can use the gpt StatisticsOp for hdf4 files.

Thanks,
Aynur

gnwiii
Posts: 605
Joined: Fri Jan 29, 2021 5:51 pm America/New_York
Answers: 1

Doubt in Statisticsop under GPT tool

by gnwiii » Thu Jan 30, 2020 8:38 am America/New_York

Have you seen the StatisticsOp Cookbook?   The examples use a .par file to provide values for parameters like band names that are not mentioned in gpt StatisticsOp -h.

OB SeaDAS - aynur
Subject Matter Expert
Subject Matter Expert
Posts: 32
Joined: Tue Dec 20, 2011 12:28 pm America/New_York

Doubt in Statisticsop under GPT tool

by OB SeaDAS - aynur » Thu Jan 30, 2020 10:13 am America/New_York

In  your XML file, please change this line    
          <source>${source}</source>
                       to 
<sourceProducts>${source}</sourceProducts>.

If you still have a problem, please send me your HDF file.

Thanks,
Aynur

ivgbhavani
Posts: 16
Joined: Mon May 21, 2018 11:23 am America/New_York

Doubt in Statisticsop under GPT tool

by ivgbhavani » Fri Jan 31, 2020 1:21 am America/New_York

Thanks Aynur, I've checked every option from my side. Its working now.
Band name (Variable name)  in hdf4 is showing two different names for my product. L_xxx in hdp dumpsds , Lt_xxx in seadas GUI. That only made error.
When I use -Pband=Lt_443 i got statistics.

Thanks for your time and support.

Post Reply