Very useful summary, many thanks. One small point, in cell 3, discussed 5-6 mins in, is it possible that A needs to be transposed i.e. np.matmul(Z,A.transpose()) - this then makes the axes in the first subplot of figure 1 look more sensible, and agrees with output I'm seeing in R when I do something similar. Apologies if I'm mistaken.
Hey - I think this is correct as is. The reason is, take a look at the second subplot in Figure 1, where X-axis is the Beta Distribution and Y is Cauchy. We know that Beta is bounded by 0-1,and the X-axis shows this. Since that is generated by the variable U_Gauss, I think it is generated correctly. That being said, I'm happy to be proven wrong if you have something more concrete? As for understanding this from Subplot-1, I think its hard to tell since that is just a joint distribution and so X and Y axis for scatter plot should look the same.
Hello, nice videos....I am very new to this and they are very helpful. I am confused about the scatterplot of figure 1 with the Gaussian Copula. Arent the inputs for all copulas between 0 and 1? Then how is the scale varying 2.5 and -2 and all in Figure 1 when Gaussian copula is imposed.
Hi there - thanks for your question. I've answered your question in this Q/A video - please have a look and let me know if you have any follow up questions. ruclips.net/video/ql4peyWtzvU/видео.html
Thanks for the lectures!
Thank you for the videos!
Thank you for the videos :)
Very useful summary, many thanks. One small point, in cell 3, discussed 5-6 mins in, is it possible that A needs to be transposed i.e. np.matmul(Z,A.transpose()) - this then makes the axes in the first subplot of figure 1 look more sensible, and agrees with output I'm seeing in R when I do something similar. Apologies if I'm mistaken.
Hey - I think this is correct as is. The reason is, take a look at the second subplot in Figure 1, where X-axis is the Beta Distribution and Y is Cauchy. We know that Beta is bounded by 0-1,and the X-axis shows this. Since that is generated by the variable U_Gauss, I think it is generated correctly.
That being said, I'm happy to be proven wrong if you have something more concrete? As for understanding this from Subplot-1, I think its hard to tell since that is just a joint distribution and so X and Y axis for scatter plot should look the same.
Hello, nice videos....I am very new to this and they are very helpful. I am confused about the scatterplot of figure 1 with the Gaussian Copula. Arent the inputs for all copulas between 0 and 1? Then how is the scale varying 2.5 and -2 and all in Figure 1 when Gaussian copula is imposed.
Hi there - thanks for your question. I've answered your question in this Q/A video - please have a look and let me know if you have any follow up questions.
ruclips.net/video/ql4peyWtzvU/видео.html
Thank you so much, nice videos!!!
thanks a lot!
very good
i d have started with the graphes 1st, then the abstract theory, you almost lost me on too many cold concepts before the answers