Use my code "SHASHANKKALANITHI" or link rize.io/shashankkalanithi to be the first 1000 people to get a 25% discount off your first three months with Rize
I have been watching u from past 6 months and ur consistently improving ur skills in TH-cam as well as in the analytic world . Thanks for everything ❤️
Hey, great video ! You definetely can use chisquared test in more than 2 categories, but your conclusion would be: the sample shows that the two variables are (or not) independent, but you wouldn't have any information on where the independence is ( a post hoc test could be useful then). Another thing as someone already mentioned is that the null hypothesis of this test is that the 2 variables are independent, so getting a p-value < 0.05 (assuming 5% significance level) means you reject H0 and concludes that you can't detect independence on this sample, so a correlation between the categories might exist! Anyways thanks for sharing everything!
Absolutely great work. I'm amazed at how well you use all your knowledge. Learning Python and Statistics doesn't make one great data analyst or scientist...but you sir, are a beast
Hello shishank kindly share with us the link to the data set ...i will like to practice with it...I stumbled on your data cleaning challenge video and that got me my first job as a data analyst ever since i have been hooked...please just drop the link to the final data set
I think you may be right. Usually it's null hypothesis > .05, reject if under, so I'm confused on his final results. However, when looking at the results of the categories themselves, I think the results look right. Maybe I'm confused because I don't do much hypothesis testing in python
The Chi-Square Test (for independence) stats H0: variables are independent. Thus for p < .05 we reject H0 which means variables are corelated. For p > .05 we accept H0 which means variables are independent. With that said usually the null hypothesis says something along: "there is no ..."; in this case, there is no relationship
With p>0.05, we assert that we failed to reject the null hypothesis which does not automatically mean we have accepted.. we didn't have sufficient evidence
To reach a different audience and expand a little bit, you could try to branch out to twitch for some of your "Ask an Analyst" live streams, I'm not sure if you will see high growth from it, but its an idea. You could also try uploading any "quick tips" you have into TH-cam shorts as I feel the algo favors shorts in some instances. I love your content, and would love to see the channel grow!
Hey Shashank thank you for your videos. It really helps me in this data analyst road. I want to ask something about Visual Code. Is there a way to show "Help Infos(Shift+Tab )" like in the jupyter notebook in VS Code? Or do you know any addons for that ? Thank you.
Use my code "SHASHANKKALANITHI" or link rize.io/shashankkalanithi to be the first 1000 people to get a 25% discount off your first three months with Rize
Holy smokes! I had no idea about the Google EDA guide 🤯 Thanks for sharing this, Shashank!!
Of course thanks for stopping by Luke! Yeah I was working on a project at work and used it for guidance so I thought I'd share.
I have been watching u from past 6 months and ur consistently improving ur skills in TH-cam as well as in the analytic world . Thanks for everything ❤️
totally agree
Thank you so much Akash! That means so much! I really tried to improve my editing skills on this one.
Hey, great video ! You definetely can use chisquared test in more than 2 categories, but your conclusion would be: the sample shows that the two variables are (or not) independent, but you wouldn't have any information on where the independence is ( a post hoc test could be useful then). Another thing as someone already mentioned is that the null hypothesis of this test is that the 2 variables are independent, so getting a p-value < 0.05 (assuming 5% significance level) means you reject H0 and concludes that you can't detect independence on this sample, so a correlation between the categories might exist! Anyways thanks for sharing everything!
Absolutely great work. I'm amazed at how well you use all your knowledge. Learning Python and Statistics doesn't make one great data analyst or scientist...but you sir, are a beast
Great video
Wonderful as usual. Thanks for everything.
And thank you so much for watching ! :)
Among the 16% women on your channel :) Love the content!
Hello shishank kindly share with us the link to the data set ...i will like to practice with it...I stumbled on your data cleaning challenge video and that got me my first job as a data analyst ever since i have been hooked...please just drop the link to the final data set
I think you mixed up the conclusion from p-value, I believe if p < 0.05 there is relation [dependant] between the variables , else not [dependant]
I think you may be right. Usually it's null hypothesis > .05, reject if under, so I'm confused on his final results. However, when looking at the results of the categories themselves, I think the results look right. Maybe I'm confused because I don't do much hypothesis testing in python
The Chi-Square Test (for independence) stats H0: variables are independent. Thus for p < .05 we reject H0 which means variables are corelated. For p > .05 we accept H0 which means variables are independent. With that said usually the null hypothesis says something along: "there is no ..."; in this case, there is no relationship
With p>0.05, we assert that we failed to reject the null hypothesis which does not automatically mean we have accepted.. we didn't have sufficient evidence
it was so good..
To reach a different audience and expand a little bit, you could try to branch out to twitch for some of your "Ask an Analyst" live streams, I'm not sure if you will see high growth from it, but its an idea. You could also try uploading any "quick tips" you have into TH-cam shorts as I feel the algo favors shorts in some instances. I love your content, and would love to see the channel grow!
Thats a good idea! It's probably the only major platform I don't have an audience on. Thanks so much for all of the support!
Thanks for the great video ! By the way, may I know how to get the data ?
Great video! Thanks 🙌🏼
Thank you so much Ricardo!
hey shashank, great video again, thank you for the content! can you provide us the notebook?
Thank you so much Leandro. Unfortunately, because the notebook has a lot of the exact URL's I access within my company I can't do that, I am sorry.
Hey Shashank thank you for your videos. It really helps me in this data analyst road.
I want to ask something about Visual Code. Is there a way to show "Help Infos(Shift+Tab )" like in the jupyter notebook in VS Code? Or do you know any addons for that ? Thank you.
Hello Shashank! Great content. Can you link the TH-cam News channel you mentioned? Thanks
Yes! Here is the specific video I referenced! Thanks so much for watching so much of the video! th-cam.com/video/404IeUzGNZ4/w-d-xo.html
Thank TH-cam very much. 😀
Is it possible for you to share the notebook file with us?
Hey Yonghai, because the notebook has a lot of the exact URL's I access within my company I can't do that, I am sorry.
@@ShashankData understood! Thanks.
Im an emory student are you an alumni?
How much foundation do I need to understand this?
I would recommend this free course in Python that I created. Thanks for stopping by! th-cam.com/video/sZDgJKI8DAM/w-d-xo.html