วีดีโอ

Welcome

It depends on theory and requirement. For symmetric weights ideally the data has same units.

You can email me at nouman.arshed@gmail.com

Thank

Thanks

Sure you can discuss at nouman.arshed at gmail.com

It is shared in previous comments

You have to do it one country at a time. There is no loop that i know of

Thanks

You can get it here drive.google.com/file/d/1AQs80qXXl-iwq4XAaYBwCG5xLCeVMz1j/view?usp=drive_link

You can fine this in any statistics book in which there is a chapter on sampling distribution and law of large numbers. Yes you can ignore normality test using this assumption. But there are some advanced methods that do not require this assumption like quantile regression models

You can access the document here - docs.google.com/document/d/11MVQIIGIdkPhCNvVXHjC7Ii7rOS0cx_6/edit?usp=drive_link&ouid=104251006931008988580&rtpof=true&sd=true

It is autometically added with zotero

you can download from here posit.co/download/rstudio-desktop/

Is there any references? My model is good, also get conintegrating relationship among variable. But I am confused regarding the statistical significance of the long-run coefficients. Please explain or refer any literature. Thanks.

Rule of thumb if t is above 2 it is significant

you can share a base paper here, i will try to develop one

you can show overall short run and add countrywise short run in the appendices

this test is checking multiple breakpoints and ranking it, it is upto you how many breakpoints you want to add in your model. this video showed how to make one dummy, in the same sheet you can write 0s and 1s for another dummy and name it dummy2

welcome

welcome

You have to install sklearn using pip install. The installation method is different for vs code, google colab and others

for that the data must be big enough to generate the MG model, in the comparison of MG and PNG you can see the crosssection wise long run shown below

you can access codes here - drive.google.com/file/d/1maBaVuoIMLrZ7xo6go_5aKJG58z41qTE/view?usp=drive_link

It would become 5

Welcome

I have not used this model yet, I will share papers if I find any

sequentially 1991 is is more significant, so if only 1 dummy is to be made i made it for 1991

Index is a normalized data where zero means mean value and negative values means that the paticular observation is less than mean.

@@nomanarshed as far I know the index should be between 0-1 or from 1-100 right? But the PCA results are not on this ranges what should I do?

It is a normalized variable not a index which ranges from - infinity to infinity

@@nomanarshed so I can use it even if it is less than 0

Yes it is providing us a linear combination of the data measing latent variable in continuous form. If its values cannot be than zero then the error term of regression should also be not zero as it is also an index as a resultant of linear combination of dv and ivs

welcome

nouman.arshed@gmail.com

write in stata webuse abdata you will get the data

you are welcome

@@nomanarshed Suppose there is raw data i.e. USD millions, and then there is data in percentages e.g. GDP growth rate, some variables as percentage of GDP, some ratios and also a few variables which are in Percentile Rank. Plz if you can tell if there is a method to transform these variables into standardized form? .... For panel data analysis

The slope coefficient adjusts the differences in the scale of variables. The best is if you can male them in relative forms like as percent of some thing so that that are comparable across countries. If you standardize them their their interpretations will be done differently

@@nomanarshed Ohk got it. Thank you so much. But I am too confused regarding specifically the Percentile Rank data. That can't be converted even!

@@DivyaSharma-jp3ki that is already transformed data, no need for firther transformation

Yes, you are right. this model is for the count outcomes specifically those where the probability of higher could is lower than the probability of the lower count.

There is no IMF code you can open the website and get the data

@@nomanarshed how can i get it.please help me

Actually there are many papers that have used it currently i dont have any that can be used as example. You can use following method to find relevant articles th-cam.com/video/c-xQ9IaF2gE/w-d-xo.html

This is output generated using boot_ardl command

Walaekum Salam. This command does not work when there are too many independent variables

there are two sources of autocorrelation, serial autocorrelation which is because of missing important variables, and pure autocorrelation which is because of data being non stationary. The second case is true when the number of years per cross section are many, in that case you may have to use panel ARDL. For the case of heteroskedasticity, you might have to explore reasons, it can be because of slope heterogeneity across cross sections for that you might have to use random coefficients models, it might be because of non linear behaviour of variable for that you might use non linear model (quadratic or moderating) or take log of variables. it may be because of outliers, then you might have to find outlier observations and remove them. Some details are provided at th-cam.com/video/vRndUgRFFmo/w-d-xo.html