Видео

You can do after you read the file to look at the data, or you can create it after you cleaned the data. It is up to you

Check the spread between quartiles: 25% (Q1), 50% (median), 75% (Q3) If a value is much higher than Q3 or much lower than Q1, it's likely an outlier. Also, if min or max values are very far from Q1 or Q3 respectively, they might be outliers To detect outliers you can use either Z-score method or IRQ method. You can find it in my Pandas Cheat Sheet stan.store/KarinaDataScientist/p/complete-pandas-cheat-sheet

@karinadatascientist someone says we can't use IQR on Skewed data, It's really true ???

thank you

Welcome!!

Thank you

Sure, I will make videos for numpy too

welcome!

thank you

Numpy and pandas go hand-in-hand. Knowing NumPy helps you understand the inner workings of pandas. Pandas is enough for data cleaning and manipulation, however numpy is required for high-performance numerical operations (because it works way faster). Pandas works with 2D data, whereas when you work with higher-dimensional data like images and tensors, you need numpy. You are already comfortable with numpy, so why not to learn pandas/polars :) ?

of course!

Thank you!

Hi Ahmed. Of course tag me on Linkedin, that's very kind of you. Thank you

@karinadatascientist Hi Karina, Thank you, I'm very happy by your response.

welcome!

😂

yes, other people heard it too! On TikTok they didn't hold it back, which made this video go sort of viral, getting nearly 23,000 views 😂.

​@@karinadatascientist don't worry about it. I really love your accent personally ❤️❤️

It is a manual process, I had to just try different options till it was placed where I wanted it to be. For example, with line labels, this is a code I used: labels = df_transformed.columns offsets = [0, 0, 0, 5, 0, 2] # Adjust positions to avoid overlap for i, (label, offset) in enumerate(zip(labels, offsets)): y_pos = df_transformed[label].dropna().iloc[-1] + offset ax.text(df_transformed.index[-1] + 2, y_pos, label, color=COLOR_SCALE[i], fontsize=10, va='center') I started with offset all zeroes, didn't like it, started adjusting one by one to see what works best. Same with x, y here: fig.text( 0.015, 0.90, # (x,y) coordinates "Share of boys or girls within the relevant age group who are enrolled in primary, secondary, and tertiary education.", color=GREY30, fontsize=12, )

edX courses are good and free. You have to try various one to see which resonates. With some you will not like instructor, with others - boring tasks. You need a combination of courses to find what you like :)

hi can u suggest any good course on generative ai

hi is coursera a good choice

Thank you 🙌

Oh, can do. Thank you for an idea

Thank you. Any business related projects - customer related, product related, expenses. Marketing campaign analytics (google, facebook, tiktok), user customer analysis (to help create a customer avatar), sales analysis. That's why I believe that sports or health projects are great, especially for practicing, but shouldn't be included in the portfolio, if you want to work for e-commerce, for example.

​@karinadatascientist Thanks! An e-commerce sales dashboard was on my list, but I wanted to do a Covid-19 Vaccination Analysis and a healthcare readmission prediction using logistic regression. Should I drop these in favor of more business oriented projects? I know I'm asking a lot of questions haha, but any help would be greatly appreciated!

@@nindogaming4132 I would avoid COVID19 data. It was very popular in 2019 and 2020 :) Choose more business oriented

@@karinadatascientist Thank you!

welcome!

You're very welcome!

Fair enough. If you know how to use index+match, you can achieve a lot in Excel!

thank you

can u suggest good books on python

hi karina can u suggests some books on datascience

​@@karthiksundaram544 o'reilly books are pretty good

is data scientist salaries good in US

I think transitioning into data engineering can be a good choice, if you are interested in that field. AI will not replace data science or data analytics, but it is another tool that will be a part of our daily workflow

Glad to hear that! thank you for the feedback

Great to hear! Will make more of them!

You're welcome 😊

You're so welcome! Thank you for watching

I will work on it!

The CV template I show is available as a part of my freebies package, that you get once you sign up to my newsletter stan.store/KarinaDataScientist/

Thank you. I hope you would enjoy the course!

thank you!!!!

Thank you! Hello, Brazil!