xgrads.core about xgrads

[miniufo]

Maybe you did not install the package. That is also fine. You can:

import sys
sys.path.append('/absolute/path/to/your/xgrads/folder')
from xgrads import open_CtlDataset

[miniufo]

Try this:

dset ^IVT0-Copy.dat
title IVT0 data
undef -9.99e33
xdef 101 linear -20 1
ydef 61 linear -10 1
zdef   1 linear 0 1
tdef   1 linear 01Jan2000 1dy
vars 1
IVT0 0 99 comment
endvars

zdef, tdef, and vars 1 cannot be missing. The rule is that, the ctl and binary data can be opened by GrADS. Sometimes, people write their own (incorrect) ctl may not correctly parsed by GrADS, and hence xgrads.

[miniufo]

xgrads is just for loading data into memory, not fully reproduce GrADS' plot functionalities. But you can use matplotlib for plotting, which is most popular in python.

[miniufo]

Of course yes. If your ctl is associated with a 4D dataset (time, lev, lat, lon), you can open it and play with 4D variables.

[miniufo]

When you load your dataset (end with .ctl) as dset, it is already a xarray.Dataset.

[miniufo]

Yes, you should import xarray as xr and call functions in it to help you. xarray is the data struct you play with, while xgrads is an input/output tool.

[FaeghPazhooh]

Thanks a lot. I omitted .core, and it worked for test files.
Now I have a binary file that is 85th percentile of 40 binary files by one variable. I want to open this file and I wrote this ctl file for it, but it does not work or plot it.
ctl:
dset IVT0-Copy.dat
title IVT0 data
undef -9.99e33
xdef 101 linear -20 1
ydef 61 linear -10 1
IVT0 0 99
endvars

from xgrads import open_CtlDataset
dset = open_CtlDataset('E:/dissertation/Code/code2/ivt/Prog/output/IVT0out/IVT0-Copy.ctl')

print all the info in ctl file

print(dset)

for your ctl content, you can plot any variables (e.g.,

first time and level of T) immediately as

dset['IVT0'][-20,-10,...].plot()

[FaeghPazhooh]

Thanks a lot for your help!
I need a 3d plot for lon, lat and the variable should display in Z. Is it available in xgrads? something like plot_surface on matplotlib.
And for comparing I should put data in a variable. Is there any method to use lon, lat and IVT in diferent times(t) in a array on xgrads?

[FaeghPazhooh]

Thanks a million !!!!
How I can use data on memory in a variable?
May I use open-ctldataset to load data in a 3d variable?

[FaeghPazhooh]

Many thanks, I opened all of my files by xgrads, and now is there any way to compare these datasets?

I have a threshold in a dataset :
dset ^IVT0-Copy.dat
title IVT0 data
undef -9.99e33
xdef 101 linear -20 1
ydef 61 linear -10 1
zdef 1 linear 0 1
tdef 1 linear 01Jan2000 1dy
vars 1
IVT0 0 99 comment
endvars

and 40 datasets as like as below:
dset E:/dissertation/Code/code2/ivt/Prog/output/IVT0out/IVT_1980.dat
title IVT data
undef -9.99e33
xdef 101 linear -20 1
ydef 61 linear -10 1
zdef 4 linear 0 1
tdef 852 linear 01Nov1979 6hr
vars 1
IVT 0 99
endvars,

I want to compare each tdef with the threshold dataset. Is there any command to compare datasets?
Thank you so much in advance, if you can introduce a method in a library to compare these kinds of datasets?

[miniufo]

I don't know what you want to do. What do you mean by compare datasets? Compare tdef with a dataset? What does that mean? If you want to compare time steps with a reference time, you can just do

print(tdef[0] > ds_threshold.tdef[0])

[FaeghPazhooh]

I want compare every tdef with a threshold(it is a separate dataset) and change the variable after comparing. I am writting From p in range(tdef) If dset1['Ivt'][p]>=dset[Ivt0][0]: dset1['Ivt'][p]==dset1['Ivt'][p] else: dset1['Ivt'][p]==0 p=p+1 but the dset1 dosenot change to new dset that I want after if.

…

On Fri, May 26, 2023, 12:02 AM Yu-Kun Qian ***@***.***> wrote: I don't know what you want to do. What do you mean by compare datasets? Compare tdef with a dataset? What does that mean? If you want to compare time steps with a reference time, you can just do print(tdef[0] > ds_threshold.tdef[0]) — Reply to this email directly, view it on GitHub <#45 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/A73ERQKBBGVKBOR7VRPR7KLXH66UJANCNFSM6AAAAAAYBIFY5I> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[miniufo]

I don't think it is possible to modify all the times of a dataset as 0, because it is a coordinate that should be monotonic increasing or decreasing. But you can assign the whole tdef with a new series of times that you are happy with.

By the way, == is not a valid assignment, should be =.

[FaeghPazhooh]

"=" is not accepted.
I have a dataset. It has just one tdef. I call it IVT0.

There is another dataset,IVT , It has 855 tdefs.

Now, in new step, I want to compare each tdef in IVT to IVT0. Grade by grade in the grid (lon and lot).

for p in range(0,855,1):

if dset1['IVT'][p,...,...].any() >= dset['IVT0'][0].any():
dset1['IVT'][p,...,...].any() == dset['IVT'][p].any();
else:
dset1['IVT'][p,...].any()==0
print('dset==0',p)
p=p+1

Overall, I want to make a new dataset that every grade value has changed after 'for', and use the new one.

[miniufo]

You can do if, and if true, assign new series of tdefs to the coordinates. You may want to search 'how to modify coordinates in xarray' in google.

[FaeghPazhooh]

At first step, I should load dataset to xarray in python. How I can open my dataset in xarray. I found "dataset to array" but it didnot work.

[miniufo]

After a quick look at previous posts, I guess you want to combine 40 datasets into a single one, right? So you have 40 ctls and want to load them all into memory?

[FaeghPazhooh]

I think I could not explain it well. I have a dataset that I want to use it as a threshold and I loaded it as "dset".

There are 40 datasets, that every one has 855 tdef. I want to compare the value of each dataset in each tdef with the value of dset and make new dataset to apply more condition.
I think I should open my file in xarray to compare and change some arraies.
How I can load my dataset in xarray?

[FaeghPazhooh]

So I should import xarray library and in the following I can use its commands for all of my datasets?

[FaeghPazhooh]

Thanks a million