Description of SCE_data not given

[vinni94]

From the code I could understand that obs and mod data are modeled and observed datasets. What is SCE_data? please provide the documentation for the same

[Vik-1]

I have the same question too, @pankajkarman could you please help to clarify this. Thanks in advance

[vinni94]

Hi Vik-1, I recently looked in to the code and realized that you can either bias correct model to observed or vice versa. so the third argument can be as same as your model or observation array. Have a great day !!

[Vik-1]

Thank you

[pankajkarman]

This package needs 3 datasets. One is reference dataset which is generally the observation mostly ground based. Second is model historical datasets which would have the same time stamp as the reference dataset. These 2 datasets are used for estimating the bias values which is instead to correct the third scenario dataset.

I hope It's clear now. BTW I'll be updating the documentation in coming weeks.

[Vik-1]

Thank you Am Donnerstag, 14. Juli 2022, 23:58:45 GMT+8 hat Pankaj Kumar @.***> Folgendes geschrieben:

This package needs 3 datasets. One is reference dataset which is generally the observation mostly ground based. Second is model historical datasets which would have the same time stamp as the reference dataset. These 2 datasets are used for estimating the bias values which is instead to correct the third scenario dataset.

I hope It's clear now. BTW I'll be updating the documentation in coming weeks.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: @.***>

[alexntougias]

I am a bit confused. The third scenario dataset are the historical data that we use as the second input? Or future scenario (e.g. ssp126) datasets?

[goharShoukat]

So in essence, the model data and the sce data can both be the same? The model data is the one that has to be corrected in my particular case.

[vinni94]

That's correct. As per my understanding.

On Mon, 15 Aug 2022, 19:37 Gohar Shoukat, @.***> wrote:

So in essence, the model data and the sce data can both be the same? The model data is the one that has to be corrected in my particular case.

— Reply to this email directly, view it on GitHub https://github.com/pankajkarman/bias_correction/issues/17#issuecomment-1215054145, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJ276AGFDT76GMOVOE5VAR3VZJFINANCNFSM5ZIPSFWA . You are receiving this because you authored the thread.Message ID: @.***>

[Anjali-Thomas95]

I too had a similar confusion. My observation data - A, and my model output (to be corrected) is B. I don't have a third set of that which is supposed to be the model historical. So what should I do in such a situation.

[vinni94]

Use 3rd argument as the model data if you need to correct to the observation.

On Wed, 17 Aug 2022, 06:19 Anjali Thomas, @.***> wrote:

I too had a similar confusion. My observation data - A, and my model output (to be corrected) is B. I don't have a third set of that which is supposed to be the model historical. So what should I do in such a situation.

— Reply to this email directly, view it on GitHub https://github.com/pankajkarman/bias_correction/issues/17#issuecomment-1217316213, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJ276ACPWGUI7NQUZHBROCLVZQZIZANCNFSM5ZIPSFWA . You are receiving this because you authored the thread.Message ID: @.***>

[Anjali-Thomas95]

Thank you for the prompt reply.

[Vik-1]

Hi @all, something is wrong with the arguement, the correct thing is to correct the model output using observation and not the other way round. Therefore, it should be made very clear. The confusion for here is; If you have model output: historical and future projection datasets (hist. and SSPs). According to the bias correction eqns (defending on a particular method). You use the observation to correct the historical and thereafter using the same for future projection eg SSP245. Having said that, a user of this code is expecting 1. bias corrected historical data and future bias corrected data as outputs. Given the above, what is the 3 variable? is it future projection data? hope we can get two ouputs? if that is the case. regards 

Sent from Yahoo Mail on Android

On Wed, Aug 17, 2022 at 8:52 AM, @.***> wrote:

Use 3rd argument as the model data if you need to correct to the observation.

On Wed, 17 Aug 2022, 06:19 Anjali Thomas, @.***> wrote:

I too had a similar confusion. My observation data - A, and my model output (to be corrected) is B. I don't have a third set of that which is supposed to be the model historical. So what should I do in such a situation.

— Reply to this email directly, view it on GitHub https://github.com/pankajkarman/bias_correction/issues/17#issuecomment-1217316213, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJ276ACPWGUI7NQUZHBROCLVZQZIZANCNFSM5ZIPSFWA . You are receiving this because you authored the thread.Message ID: @.***>

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: @.***>

[goharShoukat]

This package is quite different in its results compared to the R package. It would be interesting to see the difference between the two.