Data file download failed about imdlib

[iamsaswata]

What version of imdlib you are using?

[pradeep2c1]

I downloaded it using conda with python 3.7.9
it is: imdlib-0.1.8.1

[pradeep2c1]

Also, actually I want to work with real time data (date-wise)
But there also I got error:

[iamsaswata]

Yes, real time date (date-wise) will work if you update your imdlib to the newest version. Also, it will solve the earlier error (File Download Failed! Error: 404 Client Error) you were facing. Update the imdlib and make sure it has version 0.1.15 (as of now February, 2023). You can try something like : pip install imdlib==0.1.15 or pip install --force-reinstall -v "imdlib==0.1.15".

[pradeep2c1]

Thanks, it worked for downloading yearly data.
But still there is issue with daily data:


Can you please help out with this

[iamsaswata]

Happy to know it solved your problem.

Regarding your mentioned issue in the last comment " But still there is issue with daily data:", I can confirm you that it is not a problem of the IMDLIB library. I have attached a screenshot from my notebook for your reference.

If i have to make it simple, imd only stores real time daily data for the recent time period (that's why newer date woks, e.g., 2020 in the example), but not the data from long back (that's why 2013 in your example is not working).

[pradeep2c1]

Yes, finally I got access to real-time data after 2016.
But, I needed to work with data since 2000 or so. And that data is available at the IMD website too. But there you can only download for a day at one time, and for example, 100 days per year for so many years would be just too time consuming.
So, is there some way or solution to that?

[iamsaswata]

Happy it worked for you. If you need daily data ( which is not real time), it is possible ( take the archive data and extract as per your need). Otherwise, I am not sure. But in practice, real time data are used in real time and historical data for long term applications.

[pradeep2c1]

Maybe, my words were not that precise.
I wanted to mean that I want data like for dates 01-04-2001, 02-04-2001, ...

[iamsaswata]

That's fine.
If you can read the data in imdlib, then you can achieve this in two step. Assuming you have the long data in a variable called 'data', (1) you convert it to xarray format by using imdlib function get_xarray() function ( e.g. ds=data.get_xarray()) ( available at documentation site:https://imdlib.readthedocs.io) and finally (2) get your required days ( e.g. 01-04-2001) using xarray slice syntax ( e.g., ds.sel(time='YYYY-MM-DD')).