Forecasting Time-Series data with Prophet – Part 2

Note: There’s been some questions (and some issues with my original code). I’ve uploaded a jupyter notebook with corrected code for Part 1 and Part 2.  The notebook can be found here.

In Forecasting Time-Series data with Prophet – Part 1, I introduced Facebook’s Prophet library for time-series forecasting.   In this article, I wanted to take some time to share how I work with the data after the forecasts. Specifically, I wanted to share some tips on how I visualize the Prophet forecasts using matplotlib rather than relying on the default prophet charts (which I’m not a fan of).

Just like part 1, I’m going to be using this retail sales example csv file find on github.

For this work, we’ll need to import matplotlib and set up some basic parameters to be format our plots in a nice way (unlike the hideous default matplotlib format).

With this chunk of code, we import fbprophet, numpy, pandas and matplotlib. Additionally, since I’m working in jupyter notebook, I want to add the %matplotlib inline instruction to view the charts that are created during the session. Lastly, I set my figuresize and sytle to use the ‘ggplot’ style.

Since I’ve already described the analysis phase with Prophet, I’m not going to provide commentary on it here. You can jump back to Part 1 for a walk-through.

At this point, your data should look like this:

sample output of sales forecast

 

Now, let’s plot the output using Prophet’s built-in plotting capabilities.

 

 

Plot from fbprophet

While this is a nice chart, it is kind of ‘busy’ for me.  Additionally, I like to view my forecasts with original data first and forecasts appended to the end (this ‘might’ make sense in a minute).

First, we need to get our data combined and indexed appropriately to start plotting. We are only interested (at least for the purposes of this article) in the ‘yhat’, ‘yhat_lower’ and ‘yhat_upper’ columns from the Prophet forecasted dataset.  Note: There are much more pythonic ways to these steps, but I’m breaking them out for each of understanding.

You don’t need to delete the ‘y’and ‘index’ columns, but it makes for a cleaner dataframe.

If you ‘tail’ your dataframe, your data should look something like this:

final dataframe for visualization

You’ll notice that the ‘y_orig’ column is full of “NaN” here. This is due to the fact that there is no original data for the ‘future date’ rows.

Now, let’s take a look at how to visualize this data a bit better than the Prophet library does by default.

First, we need to get the last date in the original sales data. This will be used to split the data for plotting.

To plot our forecasted data, we’ll set up a function (for re-usability of course). This function imports a couple of extra libraries for subtracting dates (timedelta) and then sets up the function.

This function does a few simple things. It finds the 2nd to last row of original data and then creates a new set of data (predict_df) with only the ‘future data’ included. It then creates a plot with confidence bands along the predicted data.

The ploit should look something like this:

Actual Sales vs Forecasted Sales


Hopefully you’ve found some useful information here. Check back soon for Part 3 of my Forecasting Time-Series data with Prophet.

Eric D. Brown , D.Sc. has a doctorate in Information Systems with a specialization in Data Sciences, Decision Support and Knowledge Management. He writes about utilizing python for data analytics at pythondata.com and the crossroads of technology and strategy at ericbrown.com

17 thoughts on “Forecasting Time-Series data with Prophet – Part 2

  1. Hi Eric, useful post indeed. I am getting a single error though while i am plotting.
    ” NameError: name ‘predict_df’ is not defined”..Could you please help me out here ??

    Thanks a lot

    Ajeet

      1. Yes Eric, I defined it ..Pasting my complete code:

        # import library
        from datetime import date,timedelta

        #define a function
        def plot_data(func_df, end_date):
        end_date = end_date – timedelta(weeks=4)
        mask = (func_df.index > end_date) # set up a mask to pull out the predicted rows of data.
        predict_df = func_df.loc[mask]

        # Now…plot everything
        fig, ax1 = plt.subplots()
        ax1.plot(data.y_original) # original traffic data
        ax1.plot((np.exp(predict_df.yhat)), color=’green’, linestyle=’:’)
        ax1.fill_between(predict_df.index, np.exp(predict_df[‘yhat_upper’]), np.exp(predict_df[‘yhat_lower’]), alpha=0.5, color=’darkgray’)
        ax1.set_title(‘Wesbite Traffic Original (Orange) vs Wesbite Traffic Forecast (Green)’)
        ax1.set_ylabel(‘Wesbite Traffic’)
        ax1.set_xlabel(‘Date’)

        # change the legend text
        L=ax1.legend() #get the legend
        L.get_texts()[0].set_text(‘Actual Traffic’) #change the legend text for 1st plot
        L.get_texts()[1].set_text(‘Forecasted Traffic’) #change the legend text for 2nd plot

        Am i missing something ??

        And thank you so much for the prompt reply 🙂

          1. Hi Eric, thanks for the reply.

            So I assume all the commands you have written under (now..plot everything) and (change the legend) are a part of plot_df function based on indentation.

            Now it runs but neither I get any error nor the graph …Thanks

    1. Hi Eric,

      I used plt.show() but nothing appeared. Have checked multiple times. But i will fix this.
      And
      Your posts have helped me a lot doing some forecasting projects. I am a beginner in python and website has really good stuffs in python. Hope to see the 3rd part of this series soon.. And thanks again for the help 🙂

      Ajeet

  2. Hi Eric, this really helpful. I have one question, though it’s not related to forecasting. I just want to know whether a prophet model can be saved in a .pkl file like other scikit-learn model.

    1. It looks like you can dump the model into pickle and then load it and have it work. I just tried this code and it works.

      import pickle
      save_model = pickle.dumps(model)
      model2 = pickle.loads(save_model)
      model2.plot(forecast);

  3. Hi Eric,
    Thanks for a great tutorial. I am currently working on project in which I am trying to split the revenue in to seasonality and trend component.
    I had a question on daily_seasonality==true,weekly_seasonality==true and yearly_seasonality==true parameters.

    I understand that the setting daily_seasonality==true,it will fit the model for daily seasonality and the “seasonal” and “trend” output that I will get in the model output would be based on daily seasonality. Similarly for weekly_seasonality==true, the seasonal and trend component would be based on weekly and similarly for yearly.

    Is my understanding correct?

    1. I haven’t worked much with daily seasonality in the new version of prophet (v0.2) but weekly and yearly will output seasonality and trend components for those periods.

Leave a Reply

Your email address will not be published. Required fields are marked *