Python Pandas Tutorial (Part 10): Working with Dates and Time Series Data
ฝัง
- เผยแพร่เมื่อ 22 ก.ค. 2024
- In this video, we will be learning how to work with DateTime and Time Series data in Pandas.
This video is sponsored by Brilliant. Go to brilliant.org/cms to sign up for free. Be one of the first 200 people to sign up with this link and get 20% off your premium subscription.
In this Python Programming video, we will be learning several different concepts about working with DateTimes and Time Series data in Pandas. We will learn how to convert values to datetimes, how to filter by dates, how to resample our dates to do some more in-depth analysis, and more. Let's get started...
The code for this video can be found at:
bit.ly/Pandas-10
StackOverflow Survey Download Page - bit.ly/SO-Survey-Download
Datetime Formatting Codes - bit.ly/python-dt-fmt
Pandas Date Offset Codes - bit.ly/pandas-dt-fmt
✅ Support My Channel Through Patreon:
/ coreyms
✅ Become a Channel Member:
/ @coreyms
✅ One-Time Contribution Through PayPal:
goo.gl/649HFY
✅ Cryptocurrency Donations:
Bitcoin Wallet - 3MPH8oY2EAgbLVy7RBMinwcBntggi7qeG3
Ethereum Wallet - 0x151649418616068fB46C3598083817101d3bCD33
Litecoin Wallet - MPvEBY5fxGkmPQgocfJbxP6EmTo5UUXMot
✅ Corey's Public Amazon Wishlist
a.co/inIyro1
✅ Equipment I Use and Books I Recommend:
www.amazon.com/shop/coreyschafer
▶️ You Can Find Me On:
My Website - coreyms.com/
My Second Channel - / coreymschafer
Facebook - / coreymschafer
Twitter - / coreymschafer
Instagram - / coreymschafer
#Python #Pandas
Hope you all are all staying safe! In this video we'll be learning a lot about working with dates and time-series data in Pandas, and also look at doing some basic plotting. In the next video, we'll be learning how to load in data to Pandas from different (Excel, JSON, SQL, etc). Let me know if there is anything else you'd like me to cover in the Pandas series. I will likely be taking a break from this series after the next two videos are released just so I can focus on some different topics.
Hey Corey
Will you be able to cover some machine learning libraries like scikit learn,tensorflow,...?
Because matplotlib and pandas that you have covered so far is related to machine learning and we can put this libraries into practical
@@kaaaaasraa49563 I'll do my best. It's a very large topic to cover, so I've been a bit intimidated to tackle such a large course, but I know many people are wanting that, so I'll do my best to start working on something soon
Hey guys checkout dfcleaner package on pypi
Hey Corey, i was wondering if you could help do a complete EDA on one dataset ,kaggle or from anywhere in which you could cover the entire operations that we've learnt in one complete tutorial video with like a real world dataset. That could be like the end of this series video or something.
Please help with numpy
3:50 convert to datetime using to_datetime
7:45 parse dates while loading data from csv
11:20 using day_name on whole series using dt class
12:20 create column of dayname
13:20 min and max methods on datetime series
15:00 filtering by dates as string
17:20 filter by to_datetime
18:40 set date as index
19:21 filter data by just passing the date in brackets
20:05 using a slice to get specific dates data
21:00 calculating average of a slice(timeframe)
22:05 getting max value of a column on a given day
23:50 resampling(breaking down by days) a whole column into a new variable
27:00 plotting with matplotlib in pandas
28:28 resampling df with multiple columns
30:55 using agg to apply different function on different columns while resampling
Would it be possible for you to tell me how can I download ETH_1h.csv file? I found just row data on bit.ly/Pandas-10. Thanks in advance.
😊
Wow! Thk you very much! I appreciated a lot!
For all those getting the following error: *FutureWarning: The pandas.datetime class is deprecated and will be removed from pandas in a future version. Import from datetime module instead.*
Just do this:
*from datetime import datetime*
and change the *pd.datetime.strptime()* to:
*datetime.srtptime()*
and it should work the same.
I used the following
import datetime as dparser
d_parser = lambda fn : dparser.datetime.strptime(fn, '%Y-%m-%d %I-%p')
this also worked. is there any advantage of your method over mine ?
@@mrrohitnayak1 its the same thing, u just named datetime as dparser, His code would call datetime.datetime.strptime() and yours call dparser.datetime.strptime()
@@alenjose3903 any advantage in using his method over mine or vice versa
@@mrrohitnayak1 no its the same, u can name it anything you want.
Consider changing 'datetime.srtptime()' to 'datetime.strptime()'
I bless the day I came across your channel
While there are already other excellent pandas tutorials on TH-cam your videos IMHO stand out for their carefully crafted examples, clarity of explanation, information density and relevance to practical applications.
I'd really love this series to be continued and cover some of the more advanced topics.
The concept of resample is so damn crazy!!! I struggle so much in other systems with this kind of aggregation and it's easy like nuts in pandas...
Corey, you are great!
Hi Corey, thank you so much for taking the time and putting this series together. This is one of the most structured series on panda on the web. I feel confident in pandas after watching the series. We would love to see something similar on the NumPy. Please, Corey...put together a Numpy series for us.
I learn through this video in 2023 as a new entry-level data analyst and non-native English. It's still relevant and lite but packed, especially for the `.resample()` method. It saves my time compared to using `.groupby()`. In addition, we need to keep in mind indice filtering. It's better to use `.loc[]` to avoid the deprecation issue. Thank you, Corey Schafer. Your channel is such a gem!
Thanks a lot for this whole series Corey!! I've watched 3/4 other tutorials between youtube and other platforms and yours were the best by far, especially due to the fact that I'm from Italy so it was easier to follow your explanations step by step
I can't wait for your plotting with Pandas tutorials too, those are gonna be pretty useful! Have a nice day a stay safe ;)
this whole series is so beautifully explained that if one is little bit aware of python and its syntax whole series can be watched without volume .
Thank you Corey for the wonderful tutorials. No paid courses can come close to the way you present the depth of the concept. Thank you for making it free and a big thank you for keeping up the good work.
These are the most developed, best explained videos on how to use pandas - thank you Corey for putting these together!
To anyone getting the message that 'date_parser' is deprecated and will be removed in a future version...
You no longer need to pass in a function, just the date string formats in the argument date_format='%Y-%m-%d %I-%p'
example:
df = pd.read_csv('ETH_1h.csv', parse_dates=['Date'], date_format='%Y-%m-%d %I-%p')
Yesterday i was searching for any tutorial that explains dates and times in pandas and i couldn't find a good material
Thanks man
For anyone getting error in the first part of the video, this is what worked for me
d_parser = lambda x: pd.to_datetime(x, format='%Y-%m-%d %I-%p')
df = pd.read_csv('data/ETH_1h.csv', parse_dates=['Date'], date_parser=d_parser)
or pass the date format directly
df = pd.read_csv("data/ETH_1h.csv",parse_dates=['Date'],date_format = "%Y-%m-%d %I-%p")
@@rafabws for dfx.loc['2020-01':'2020-02'] , I get KeyError: 'Value based partial slicing on non-monotonic DatetimeIndexes with non-existing keys is not allowed.; What to do with it? plz help
@@user-testo21347 I was getting same error too and I believe I found the solution. For some reason our indexes are not in order. For this reason, if we sort our indexes by writing the code dfx.sort_index(inplace=True) before writing the code dfx.loc['2020-01':'2020-02'], we get rid of the error.
@@user-testo21347 I was getting same error too and I believe I solved it. For some reason our indexes are not in order. For this reason, if we sort our indexes by writing the code dfx.sort_index(inplace=True) before writing the code dfx.loc['2020-01':'2020-02'], we get rid of the error.
Also to get the mean for resampling, the default for the mean function was changed. So to get the weekly resampling at around 30:05 of the video:
df.resample('W').mean() will give an error
do this instead:
df.resample('W').mean(numeric_only=True)
where to download this time series data please
Hi Corey..I have referred to various resources to learn Pandas properly on TH-cam. And your tutorials are the best I could find. And looking forward to learn more from your matplotlib tutorials. Thanks for creating top class content!!
This is the only channel I can’t wait for the new video to drop. Two weeks and waiting but who’s counting. ☺️
I really appreciate your detailed explanation on the errors, which help us to select the right corrective action to take. Love your videos.
Very, very, very good. The accurate and straightforward explanations are supported by great examples. Big thanks.
Right now I'm learning flask from your playlist and it's freaking amazing!
You taught me pandas completely free of charge.
Thank you, sir.
I'm happy that I found a Awesome channel to learn Python for Data Science Easily!!!! Very Nice and Crystal Clear Explanation!!👌🙂
Gave a thumbs up before even started watching the tutorial..and I'm not disappointed at all. Keep up the good work Corey!!!
BTW you are the only TH-camr getting likes from me.
I want to point out something. I like your videos. There is nothing boring about them. But I do not have much time so I watch them at 2x. Your normal talking speed is not slow but very clear so I can easily understand you at 2x too.
My man Corey on youtube dropping bangers.
Thanks man. Amazing content.
Your every video is soo nice and perfect. The explanation is very friendly and in a proper sequence. Your channel works for me as a one-stop-shop. Thank you soo much man.❤️❤️
Love from India 🇮🇳
Please show more *errors and its corrections* for future vids.
The fixes are really beneficial!
Thanks for recording, editing, uploading and sharing! 👍
Since I stumbled on Corey Schafer videos, it has become my favourite TH-cam Channel
great job Corey, I was going to go through Wes McKinney's book but looking at the pandas section I realized that I already know most of it by now form your series. Pls do the data plotting soon! ;)
Fantastic series! Thank you for providing this awesome content. I have been using Pandas quite a bit lately to process data, one thing I am struggling with and would really find helpful would be a video on taking these Pandas notebooks and functions and converting them into actual Python apps that can be reused on other projects. Pandas can really be like magic sometimes and being able to pass data into a Python app that could apply these functions would be extremely powerful, I have yet to find any good tutorials or reading on this subject, one from the master would be greatly appreciated!!
All I can say you are awesome. The way you explained 😍✌️✌️✌️✌️. You have the best play list for pandas on web. Your efforts are truly appreciated 👍👍👍👍👍👍
I learnt so much from this single video. Thank you Corey
I used this video to backtrack over some material that I wasn't sure of. This was a great video as is your other instructional videos. Thanks so much.
Fantastic series on Pandas. I have no words to express how i am thankful to you for this series. You are a great teacher. I am yet to check on the 11th video on this series and up next to the Matplotlib series (definetely). You are awesome. A request, could you please plan to do Numpy series as well.
Well done! I didn't know that I can pass in a dict in agg. Always passed in a list and then selected the column-aggregation tuples that I needed, which is quite resource intensive and hard to read. Also I did not know I could make datetime comparisons to a string. Thank you.
I always give you a thumbs up before i even watch the video!
Will you introduce us to machine learning in the near future?
You are a great teacher, many thanks for putting in all the effort and time in creating such a great series :)
This was an absolute masterpiece!! Thank you so much. I hope you are good and safe
pandas plotting will be useful and intesting! looking forward to it! Thanks Corey!
Extremely useful. Thanks for the video Corey! You have my gratitude.
I really like how you explain things so clearly. Super helpful for my project. Thank you so much!
This is the best episode from the series, well done
Hi Corey! Your series on pandas is really informative and helpful. Can you make a similar series for numpy? I believe that's a topic that will be of interest to many of your viewers.
That's great Corey !!
The best lesson about Time series, ever.
Thank you Sir for sacrificing your precious time to teach the World.
We appreciate you .
l Appreciate You.
Another great video Corey. Keep up the good work!!!
As usual your videos are great ! You explain things very clearly. Please start videos on Machine Learning and Artificial Intelligence soon. Earnest request !
There should be a mega-like button for this. Thank you so much!
OMG!!!
You are awesome Corey.
I learned a lot from your videos!
I got an error with the lambda function trying to parse the dates while loading from csv. The error stated pandas no longer supports pd.datetime. After checking the comments I tried @larc99's suggestion but still got an error as it was expecting a string rather than a function for the format. What gave the same output and I think is much simpler (without the lambda function) turned out to be:
df = pd.read_csv('data/ETH_1h.csv', parse_dates=['Date'], date_format='%Y-%m-%d %I-%p')
Very well-explained. Thanks for this. I am requesting you to make a video on multi index in Pandas.
You're my savior I was searching for video as this thank you.
This Video was sent by the Data Science Gods. Thank you so much
I watched this video when the likes for this video are 999 and I hit like right away to make it 1K, because I believe Corey's content!
Corey this series is fantastic, thank you!
Hey Corey! All your teachings are really good. I just need to know where you refer for your tutorials. Apart from the documentation, can you please tell me some other references that I could use?
Pandas is so easy to work on after seeing this vid.
Wish my coaching teacher taught me this simpler way
Hi Corey, Thank you so much for the videos! I was wondering when you are going to come out with the Pandas plotting series you mention? Thanks again!
Oh Corey, you legend! Thanks mate, this is some great stuff.
I loved the point Corey made about how much easier it is to use date as your index, however while following along I got a warning this warning:
FutureWarning: Indexing a DataFrame with a datetimelike index using a single string to slice the rows, like `frame[string]`, is deprecated and will be removed in a future version. Use `frame.loc[string]` instead.
Great job Corey! Thank you!
Hy, Corey. First of all, I would like to thank you for the videos. They are very clear and comprehensive. I already became a member to help you. I have one question regarding the week agregation (around 32'.) The open and close prices for a week should be the prices for the first and last hours of the week respectively. How do I do that?? Should I create a custom function for this?? Thks, GB
just 11 minutes in the video, i hit the like button as hard as i can !!
Your tutorials are the bests on the web. Before I can do the index filter (19:35 in the video), I had to sort the dataframe by index(df = df.sort_index()) I don't know why but it was given me an error (AssertionError: ).
Thanks for sharing.
Best tutorial series for pandas ever❤❤❤
Thanks a lot for this pandas series Corey ♥️.
again and again awesome information
Cool stuff. Could you show how to do this if you have more than one "index column"? For this particular Domain, how do I aggregate over intervals by grouping variable like separate stocks
Thank you thank you thank you, Finally found the help that i needed!
Corey, Thank you for the nice vide series. Could you please provide a video on multiple indexes? Also, do you have any plans for other Python libraries? such as Numpy, Sklearn?
Regardless, Thank you so much for nice videos.
Thanks Corey!
Your videos are always lit!
😁✌️
What a detailed video !! Thank you man !!! 🤝
At 20.08, calling df['2019'] didn't work for me. Gave an 'Assertion Error'. df.loc['2019'] worked though.
Same here !
@CoreySchafer, you're awesome. Loved your quality video content. Your work is fabulous.
I'd really like to see DevOps related content if it comes under your scope. Thanks 👍
Hello, this was a great!
I was wondering wouldn't it make more sense to take the close of the last day of the week rather than mean (in the resampling part).
Any answer would be appreciated!
Thanks!
corey man, i feel obligated to say thank you , aaaaaaaaaaaaaaaaaand a big well done sir.
How to extract the Close value for each Day? Is there an Agg function for the last value similar to min/max/mean?
Hi Corey, first I`ll like to thank you for this excellent material, it has been a great series. But I do have a question, i`m currently working in a project where I need to set the value of a variable equal to the value of a cell in excel. For example, var1 = Value in cell F12, so i can use the value of this variable to operate over a data frame.
The problem has been that the cells to be collected as values are not in a data frame format, just spread across different sheets.
Thanks
Loving this series and have learned a lot. The prices of ETH in this video, wow!
டேmn true. $200 vs $4900 at peak and $1700 now
Thanks, Corey. A great series of video lessons!
I've been using .agg along the lines of:
dfGroup1 = dfTest.groupby(
["area", "instrument"]).agg(
Sales=('quantity', 'sum')
)
This allows you to rename columns. Is the dictionary method the preferred way?
Hey Corey, Shall we wait for NumPy series as well after this ?
Thank you for the detailed explanation
Great! I love your lectures.
This is so in depth I love it thank you!
After using d_pars = lambda x: datetime.strptime(x, '%Y-%m-%d %I-%p') i got this warning: FutureWarning: The pandas.datetime class is deprecated and will be removed from pandas in a future version. Import from datetime module instead. I mean, i still get correct output , but if u wanna remove this warning just do: from datetime import datetime then d_pars = lambda x: datetime.strptime(x, '%Y-%m-%d %I-%p').
Treat it like curiosity or sth
Btw, great videos, sir!
Values the time of viewers. Not sure if you really speak so fast or just increase speed in the first and last few mins of video.
I just like the way it goes. Thanks for the great tutorial.
Thank you again Corey!!
This is great. Can you also make a video on how to work with timeseries data where I also want index by another categorical column and how to work on grouping them and doing forward filling
Unrelated question: what theme do you use in you videos when using sublime text? I'm dying to know.
Amazing tutorials man ! thank you so much
Hi Corey,
I am facing the following error while loading the csv in the dataframe variable df.
ParserError: Expected 1 fields in line 278, saw 2
Can you please provide some guidance for the same?? Thank you
On executing the part of d_parser while reading the CSV file, the interpreter shows "FutureWarning: The pandas.datetime class is deprecated and will be removed from pandas in a future version. Import from datetime module instead."
How to prevent from getting any error in the future version of pandas?
This will be the video I am looking for. I am pretty sure. LET IT BE!
Pretty mad how much I failed at plotting panda dataframe in matplotlib sorted by month.
We have a german saying, „taking a detour increases the knowledge of your surroundings“.
From 0% to 5% domain knowledge in multiple hours:/
Thank god I find ur channel it’s been helpful, but I have a question.. If I had 2 columns in a data frame and I want to calculate the end_date and start_date how that can be done ?
Hi what if I have a dataset with the same date in like 6 rows but a different hour attached to each row and I want to condense the 6 date rows into just two rows of the same date and use that to group my other non-date columns into things like .agg{ revenue: 'sum', orders: 'count"}? In your dataset for example, what if I want to aggregate the dates for every day into two rows of 12 hours each? How do I do that?
Hi Corey, thanks for the video. I do have one question. When I call by df ['2019'], it will give me error: AssertionError: . I have checked the data type, the index is DateTimeIndex. Can you help to this? Thank you!
Awesome stuff as always
Thank you for the great videos!
Is it possible to resample textbody like comments? and what would be the argument for that in the resample function ?
You are the best teacher. period