Time Series Forecasting Using FB Prophet| Complete Python Tutorial|

In this video I show you how to do timer series prediction and forecasting using the facebook prophet library in python for complete beginners.
The library allows you to do time series analysis by giving you weekly and yearly components breakdown of any time series as well.

Recommended Books to get better at Time Series Analysis and Python:

If you found this helpful, do like the video and subscribe to the channel!

[namithasreenath]

You've taught me more data science in the last two hours than I have struggled to learn in the last two weeks. You deserve so many more subs and exposure! Thank you so much for these videos, you are awesome!

[saketsrivastava]

It is probably the best and short video for learning time series with fb prophet. 👍🏻

[varushsimha]

Easy to understand explanation. Cheers mate!

[rashibhardwaj]

Very easy to understand, Thank you for the video.

[ahmetbeskardes]

Hello, in which python version can we run the prophet library? I'm developing on Anaconda-Spyder but I get different errors each time.

[alx]

Hey. Nice videos, thanks for them. This particular one sadly missing visual comparison of predictions vs test dataset

[shilpatiwari-ey]

Can we select training and testing data randomly or since it is a time series data, we have to split training and testing in chronological order ?

[bhabanisankar]

Excellent video. Make some videos using hybrid models

[victorhenostroza]

Hello, Can u make another video with prophet but adding regressors? for example, predict AvgTemp and take MinTemp, MaxTemp, Sunrise and Sunset as regressors.

[gautamjaiswal]

Hello Bhaiya,
I have to ask one question related to Kaggle.
I have made a linear regression model and what I have done is that I have performed StandardScaler in it (On both dependent and Independent feature)
Now when i have used my test.csv file then in there my output is also scaled down.
So I want to know that whether should I submit this scaled down values or is there any other method ??

[trantuanthach]

plot_plotly(m, forecast)
plot_components_plotly(m, forecast)
how can show figures as in the PROPHET VISUALIZATION?

[SC-hpdn]

Thanks a lot mate! This was really helpful. Just one question, will user defined MAPE function work here? I am getting nan whenever I am trying to calculate MAPE

[sameerulhaq]

Hi but when I try installing fbprophet in jupyter it doesnt install and give me this error
Failed to build fbprophet pystan
IOPub data rate exceeded.
The Jupyter server will temporarily stop sending output
to the client in order to avoid crashing it.
To change this limit, set the config variable


Current values:
(bytes/sec)
(secs)

[djwvwkx]

Cool thank you! But what about the next year that hasnt happened?

[Gabi_]

No module named 'fbprophet' when trying to execute.

[FindMultiBagger]

Hey really nice tutorial,

I have sales data and want to predict next quarter sales.

How to consider multivariate timeseris with quarter here ?

[hunterxhunter]

lemme ask you one thing, why don't you use MAE as an evaluation metric, if you have for example outliers in your dataset RMSE wouldn't be a good choice in my opinion .

[ankusharora]

I am working on a sensor data and it has temperature values at every second of a day for a week. Can i use prophet for predicting the temperature values at every second?

[BhuvaneshSrivastava]

Nachiketa: One question.. can we use this to detect Anomalies? If yes then how?

[Dogcat]

Hi is there a way to continuously input data and predict the future values using this model ? Thx