Changes to McIDAS-V nightly builds

New McIDAS-V features under development. Feedback is appreciated.
User avatar
bobc
Posts: 933
Joined: Mon Nov 15, 2010 5:57 pm

Re: Changes to McIDAS-V nightly builds

Post by bobc »

Hi Joleen -

For Scenario B, assuming you're using the goes14.ssec.wisc.edu server, there are approximately 3,443 times on the server, so it takes a while to go through all of the times. I went back to version 1.4 and saw the same slowness that you noted. This may be related to Inquiry 1672, where when trying to display ~500 local AREAS, it was taking about 10 minutes to be able to click Create Display. We are currently in the process of integrating the IDV's image chooser, and I will test the data from the goes14 server out there to see if there is any improvement.

I also heard that you were wondering if there was a way to select a range of times to display without having to go to the Absolute times tab and wait for it to be populated. Right now, though the GUI there's no way of doing this. For adding this functionality, I wrote up Inquiry 1798.

Here are a couple of alternative ways to work with this SRSO data a little faster:
1. Try geoarc.ssec.wisc.edu/GOES14/OTHER. With the archive server, you select a date and then the Absolute times tab lists out the times just for that day. Even though this is an archive server, the data on it is pretty recent if you choose the current day (within 5 minutes of real-time). I ran a quick test here with data from 2014-08-18 18:48-18:50 and it seems to work much faster. I'm not sure why this would be, but I'll investigate and write up an inquiry.
2. If you want to use goes14.ssec.wisc.edu, I found that running a script works faster than going through the GUI. The listADDEImageTimes step takes a couple of minutes to complete. Here's a sample script you can use that will create a loop of images from 2014-08-18 18:48-18:50. You would just need to plug your accounting info into the srsoParms dictionary:

Code: Select all

srsoLoop=[]

# defining day and start/ending time for loop
day='2014230'
startTime='18:48'
endTime='18:50'

# setting up generic dictionary to pass to lister and getter functions
srsoParms = dict(
    accounting=('xxx','xxxx'),
    server='goes14.ssec.wisc.edu',
    dataset='GOES14',
    descriptor='OTHER',
    band=1,
    day=day,
    )

# listing available times from server
dateTimeList = listADDEImageTimes(time=(startTime,endTime),position='ALL',**srsoParms)
print dateTimeList

# pulling in times from server to create a loop
for dateTime in dateTimeList:
    imageTime = dateTime['time']
    metadata,data=getADDEImage(time=(imageTime,imageTime),**srsoParms)
    srsoLoop.append(data)

# build window and display loop
panel=buildWindow()
srsoLayer=panel[0].createLayer('Image Sequence Display',srsoLoop)

I'll take a look at Scenario A with the NPP data to see if I can replicate the problem.

Thanks for posting your observations and suggestions -
Bob
User avatar
bobc
Posts: 933
Joined: Mon Nov 15, 2010 5:57 pm

Re: Changes to McIDAS-V nightly builds

Post by bobc »

Hi Joleen -

I'm working on trying to reproduce your VIIRS sandwich problem, and I just wanted to make sure that I was creating my display the same way you are. Basically, after passing both bands through swathToGrid, I'm just displaying SVI01 Reflectance with a gray scale and then overlaying the SVI05 brightness temperatures masked to only show values less than 225 degrees K. Is this how you are doing your sandwich display, or are you doing something different?

Thanks -
Bob
Post Reply