View Jupyter notebook on the GitHub.

Backtest: validation on historical data#

Binder

This notebook contains the simple examples of time series validation using backtest module of ETNA library.

Table of contents

  • What is backtest and how it works

  • How to run a validation

  • Backtest with fold masks

  • Validation visualisation

  • Metrics visualisation

[1]:
!pip install "etna[prophet]" -q
[2]:
import warnings

warnings.filterwarnings("ignore")
[3]:
import numpy as np
import pandas as pd

from etna.analysis import plot_backtest
from etna.datasets.tsdataset import TSDataset
from etna.metrics import MAE
from etna.metrics import MSE
from etna.metrics import SMAPE
from etna.models import ProphetModel
from etna.pipeline import Pipeline

1. What is backtest and how it works#

Backtest is a predictions and validation pipeline build on historical data to make a legitimate retrotest of your model.

How does it work?

When constructing a forecast using Models and further evaluating the prediction metrics, we measure the quality at one time interval, designated as test.

Backtest allows you to simulate how the model would work in the past:

  • selects a period of time in the past

  • builds a model using the selected interval as a training sample

  • predicts the value on the test interval and calculates metrics.

The image shows a plot of the backtest pipeline with n_folds = 3.

backtest-scheme

Below we will call a fold the train + test pair, for which training and forecasting is performed.

[5]:
df = pd.read_csv("./data/example_dataset.csv")
df["timestamp"] = pd.to_datetime(df["timestamp"])
df = df.loc[df.segment == "segment_a"]
df.head()
[5]:
timestamp segment target
0 2019-01-01 segment_a 170
1 2019-01-02 segment_a 243
2 2019-01-03 segment_a 267
3 2019-01-04 segment_a 287
4 2019-01-05 segment_a 279

Our library works with the special data structure TSDataset. So, before starting the EDA, we need to convert the DataFrame into TSDataset.

[6]:
ts = TSDataset(df, freq="D")

Let’s get a look on series

[7]:
ts.plot(segments=["segment_a"])
../_images/tutorials_102-backtest_13_0.png

2. How to run a validation#

For an easy start let’s create a Prophet model

[8]:
horizon = 31  # Set the horizon for predictions
model = ProphetModel()  # Create a model
transforms = []  # A list of transforms -  we will not use any of them

2.1 Pipeline#

Now let’s create an instance of Pipeline.

[9]:
pipeline = Pipeline(model=model, transforms=transforms, horizon=horizon)

We are going to run backtest method for it. As a result, three dataframes will be returned:

  • dataframe with metrics for each fold and each segment,

  • dataframe with predictions,

  • dataframe with information about folds.

[10]:
metrics_df, forecast_df, fold_info_df = pipeline.backtest(ts=ts, metrics=[MAE(), MSE(), SMAPE()])
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
16:56:08 - cmdstanpy - INFO - Chain [1] start processing
16:56:08 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.1s remaining:    0.0s
16:56:08 - cmdstanpy - INFO - Chain [1] start processing
16:56:08 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.2s remaining:    0.0s
16:56:08 - cmdstanpy - INFO - Chain [1] start processing
16:56:08 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.2s remaining:    0.0s
16:56:08 - cmdstanpy - INFO - Chain [1] start processing
16:56:08 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   4 out of   4 | elapsed:    0.3s remaining:    0.0s
16:56:08 - cmdstanpy - INFO - Chain [1] start processing
16:56:08 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    0.4s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    0.4s finished
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.2s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   4 out of   4 | elapsed:    0.3s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    0.3s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    0.3s finished
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.0s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.0s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   4 out of   4 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    0.1s finished
[11]:
metrics_df.head()
[11]:
segment MAE MSE SMAPE fold_number
0 segment_a 18.784294 591.542237 4.356715 0
0 segment_a 21.025083 673.133356 4.842450 1
0 segment_a 29.946378 1301.956779 6.267483 2
0 segment_a 36.310379 2032.190048 7.493394 3
0 segment_a 31.438120 1839.589752 6.364278 4
[12]:
forecast_df.head()
[12]:
segment segment_a
feature fold_number target
timestamp
2019-06-29 0 395.126752
2019-06-30 0 386.050787
2019-07-01 0 492.965980
2019-07-02 0 509.348749
2019-07-03 0 497.527370
[13]:
fold_info_df.head()
[13]:
train_start_time train_end_time test_start_time test_end_time fold_number
0 2019-01-01 2019-06-28 2019-06-29 2019-07-29 0
1 2019-01-01 2019-07-29 2019-07-30 2019-08-29 1
2 2019-01-01 2019-08-29 2019-08-30 2019-09-29 2
3 2019-01-01 2019-09-29 2019-09-30 2019-10-30 3
4 2019-01-01 2019-10-30 2019-10-31 2019-11-30 4

You can additionally get the metrics averaged over folds:

[14]:
metrics_df, forecast_df, fold_info_df = pipeline.backtest(
    ts=ts, metrics=[MAE(), MSE(), SMAPE()], aggregate_metrics=True
)
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
16:56:09 - cmdstanpy - INFO - Chain [1] start processing
16:56:09 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.1s remaining:    0.0s
16:56:09 - cmdstanpy - INFO - Chain [1] start processing
16:56:09 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.1s remaining:    0.0s
16:56:09 - cmdstanpy - INFO - Chain [1] start processing
16:56:09 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.2s remaining:    0.0s
16:56:09 - cmdstanpy - INFO - Chain [1] start processing
16:56:09 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   4 out of   4 | elapsed:    0.3s remaining:    0.0s
16:56:09 - cmdstanpy - INFO - Chain [1] start processing
16:56:09 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    0.4s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    0.4s finished
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.2s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.2s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   4 out of   4 | elapsed:    0.3s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    0.4s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    0.4s finished
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.0s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   4 out of   4 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    0.1s finished
[15]:
metrics_df.head()
[15]:
segment MAE MSE SMAPE
0 segment_a 27.500851 1287.682434 5.864864
[16]:
forecast_df.head()
[16]:
segment segment_a
feature fold_number target
timestamp
2019-06-29 0 395.126752
2019-06-30 0 386.050787
2019-07-01 0 492.965980
2019-07-02 0 509.348749
2019-07-03 0 497.527370
[17]:
fold_info_df.head()
[17]:
train_start_time train_end_time test_start_time test_end_time fold_number
0 2019-01-01 2019-06-28 2019-06-29 2019-07-29 0
1 2019-01-01 2019-07-29 2019-07-30 2019-08-29 1
2 2019-01-01 2019-08-29 2019-08-30 2019-09-29 2
3 2019-01-01 2019-09-29 2019-09-30 2019-10-30 3
4 2019-01-01 2019-10-30 2019-10-31 2019-11-30 4

3. Backtest with fold masks#

You can use your backtest strategies using FoldMask. In order to make a FoldMask, you need to specify first_train_timestamp, last_train_timestamp, target_timestamps. Then you need to pass a list of masks as the n_folds parameter to the pipeline.backtest.

Consider 3 strategies similar to SlidingWindowSplitter, ExpandingWindowSplitter and SingleWindowSplitter from sktime.

3.1 ExpandingWindowSplitter#

For this backtest strategy you can use our backtest with mode="expand".

[18]:
metrics_df, _, _ = pipeline.backtest(ts=ts, metrics=[MAE(), MSE(), SMAPE()], n_folds=3, mode="expand")
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
16:56:10 - cmdstanpy - INFO - Chain [1] start processing
16:56:10 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.1s remaining:    0.0s
16:56:10 - cmdstanpy - INFO - Chain [1] start processing
16:56:10 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.2s remaining:    0.0s
16:56:10 - cmdstanpy - INFO - Chain [1] start processing
16:56:10 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.3s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.3s finished
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.2s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.2s finished
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.0s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.1s finished
[19]:
metrics_df
[19]:
segment MAE MSE SMAPE fold_number
0 segment_a 29.946378 1301.956779 6.267483 0
0 segment_a 36.310379 2032.190048 7.493394 1
0 segment_a 31.438120 1839.589752 6.364278 2

3.2 SingleWindowSplitter#

For this backtest strategy you can use our backtest with parameter n_folds=1 or if you need specific window from your dataset you can create a mask with FoldMask.

[20]:
from etna.pipeline import FoldMask
[21]:
# 1 Without mask

metrics_df, _, _ = pipeline.backtest(ts=ts, metrics=[MAE(), MSE(), SMAPE()], n_folds=1)
metrics_df
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
16:56:10 - cmdstanpy - INFO - Chain [1] start processing
16:56:10 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.1s finished
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.1s finished
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.0s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.0s finished
[21]:
segment MAE MSE SMAPE fold_number
0 segment_a 31.43812 1839.589752 6.364278 0
[22]:
# 2 With specific mask
window_size = 85
first_train_timestamp = ts.index.min() + np.timedelta64(100, "D")
last_train_timestamp = first_train_timestamp + np.timedelta64(window_size, "D")
target_timestamps = pd.date_range(start=last_train_timestamp + np.timedelta64(1, "D"), periods=horizon)
mask = FoldMask(
    first_train_timestamp=first_train_timestamp,
    last_train_timestamp=last_train_timestamp,
    target_timestamps=target_timestamps,
)
[23]:
# 1 Without mask

metrics_df, _, _ = pipeline.backtest(ts=ts, metrics=[MAE(), MSE(), SMAPE()], n_folds=[mask])
metrics_df
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
16:56:10 - cmdstanpy - INFO - Chain [1] start processing
16:56:11 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.3s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.3s finished
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.1s finished
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.0s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.0s finished
[23]:
segment MAE MSE SMAPE fold_number
0 segment_a 39.782294 2191.914481 9.056343 0

3.3 SlidingWindowSplitter#

To implement this backtest strategy you need to create a list of fold masks with FoldMask.

[24]:
n_folds = 3
[25]:
def sliding_window_masks(window_size, n_folds):
    masks = []
    for n in range(n_folds):
        first_train_timestamp = ts.index.min() + np.timedelta64(100, "D") + np.timedelta64(n, "D")
        last_train_timestamp = first_train_timestamp + np.timedelta64(window_size, "D")
        target_timestamps = pd.date_range(start=last_train_timestamp + np.timedelta64(1, "D"), periods=horizon)
        mask = FoldMask(
            first_train_timestamp=first_train_timestamp,
            last_train_timestamp=last_train_timestamp,
            target_timestamps=target_timestamps,
        )
        masks.append(mask)
    return masks
[26]:
masks = sliding_window_masks(window_size=window_size, n_folds=n_folds)
metrics_df, _, _ = pipeline.backtest(ts=ts, metrics=[MAE(), MSE(), SMAPE()], n_folds=masks)
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
16:56:11 - cmdstanpy - INFO - Chain [1] start processing
16:56:11 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.3s remaining:    0.0s
16:56:11 - cmdstanpy - INFO - Chain [1] start processing
16:56:11 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.6s remaining:    0.0s
16:56:11 - cmdstanpy - INFO - Chain [1] start processing
16:56:12 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    1.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    1.1s finished
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.2s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.2s finished
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.0s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.1s finished
[27]:
metrics_df
[27]:
segment MAE MSE SMAPE fold_number
0 segment_a 39.782294 2191.914481 9.056343 0
0 segment_a 40.293273 2287.327209 9.108208 1
0 segment_a 39.861472 2192.540731 9.010115 2

4. Validation visualisation#

[28]:
plot_backtest(forecast_df, ts)
../_images/tutorials_102-backtest_48_0.png

To visualize the train part, you can specify the history_len parameter.

[29]:
plot_backtest(forecast_df, ts, history_len=70)
../_images/tutorials_102-backtest_50_0.png

5. Metrics visualization#

In this section we will analyze the backtest results from the different point of views.

[30]:
from etna.analysis import metric_per_segment_distribution_plot
from etna.analysis import plot_metric_per_segment
from etna.analysis import plot_residuals
from etna.analysis import prediction_actual_scatter_plot
[31]:
df = pd.read_csv("./data/example_dataset.csv")
df["timestamp"] = pd.to_datetime(df["timestamp"])
ts_all = TSDataset(df, freq="D")
[32]:
metrics_df, forecast_df, fold_info_df = pipeline.backtest(ts=ts_all, metrics=[MAE(), MSE(), SMAPE()])
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
16:56:13 - cmdstanpy - INFO - Chain [1] start processing
16:56:14 - cmdstanpy - INFO - Chain [1] done processing
16:56:14 - cmdstanpy - INFO - Chain [1] start processing
16:56:14 - cmdstanpy - INFO - Chain [1] done processing
16:56:14 - cmdstanpy - INFO - Chain [1] start processing
16:56:14 - cmdstanpy - INFO - Chain [1] done processing
16:56:14 - cmdstanpy - INFO - Chain [1] start processing
16:56:14 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.3s remaining:    0.0s
16:56:14 - cmdstanpy - INFO - Chain [1] start processing
16:56:14 - cmdstanpy - INFO - Chain [1] done processing
16:56:14 - cmdstanpy - INFO - Chain [1] start processing
16:56:14 - cmdstanpy - INFO - Chain [1] done processing
16:56:14 - cmdstanpy - INFO - Chain [1] start processing
16:56:14 - cmdstanpy - INFO - Chain [1] done processing
16:56:14 - cmdstanpy - INFO - Chain [1] start processing
16:56:14 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.6s remaining:    0.0s
16:56:14 - cmdstanpy - INFO - Chain [1] start processing
16:56:14 - cmdstanpy - INFO - Chain [1] done processing
16:56:14 - cmdstanpy - INFO - Chain [1] start processing
16:56:14 - cmdstanpy - INFO - Chain [1] done processing
16:56:14 - cmdstanpy - INFO - Chain [1] start processing
16:56:14 - cmdstanpy - INFO - Chain [1] done processing
16:56:14 - cmdstanpy - INFO - Chain [1] start processing
16:56:14 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.9s remaining:    0.0s
16:56:14 - cmdstanpy - INFO - Chain [1] start processing
16:56:14 - cmdstanpy - INFO - Chain [1] done processing
16:56:14 - cmdstanpy - INFO - Chain [1] start processing
16:56:14 - cmdstanpy - INFO - Chain [1] done processing
16:56:14 - cmdstanpy - INFO - Chain [1] start processing
16:56:15 - cmdstanpy - INFO - Chain [1] done processing
16:56:15 - cmdstanpy - INFO - Chain [1] start processing
16:56:15 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   4 out of   4 | elapsed:    1.2s remaining:    0.0s
16:56:15 - cmdstanpy - INFO - Chain [1] start processing
16:56:15 - cmdstanpy - INFO - Chain [1] done processing
16:56:15 - cmdstanpy - INFO - Chain [1] start processing
16:56:15 - cmdstanpy - INFO - Chain [1] done processing
16:56:15 - cmdstanpy - INFO - Chain [1] start processing
16:56:15 - cmdstanpy - INFO - Chain [1] done processing
16:56:15 - cmdstanpy - INFO - Chain [1] start processing
16:56:15 - cmdstanpy - INFO - Chain [1] done processing
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    1.6s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    1.6s finished
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.2s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.4s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.6s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   4 out of   4 | elapsed:    0.8s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    1.0s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    1.0s finished
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.0s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   4 out of   4 | elapsed:    0.2s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    0.2s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    0.2s finished

Let’s look at the distribution of the SMAPE metric by folds. You can set type_plot as box, violin or hist.

[33]:
metric_per_segment_distribution_plot(metrics_df=metrics_df, metric_name="SMAPE", plot_type="box")
../_images/tutorials_102-backtest_57_0.png

Let’s look at the SMAPE metric by segments

[34]:
plot_metric_per_segment(metrics_df=metrics_df, metric_name="SMAPE", ascending=True)
../_images/tutorials_102-backtest_59_0.png

Now let’s look at the residuals of the model predictions from the backtest. Analysis of the residuals can help establish a dependency in the data that our model was not able to find. This way we can add features or improve the model or make sure that there is no dependency in the residuals. Also, you can visualize the residuals not only by timestamp but by any feature.

[35]:
plot_residuals(forecast_df=forecast_df, ts=ts_all)
../_images/tutorials_102-backtest_61_0.png
[36]:
prediction_actual_scatter_plot(forecast_df=forecast_df, ts=ts_all)
../_images/tutorials_102-backtest_62_0.png

That’s all for this notebook. More features you can find in our documentation!