'DartsForecaster' object has no attribute '_get_model_container'

[antara-c]

Hi, I am trying to run conformal tights on my local machine, however even after a pip install conformal tights and pip install darts, I get the above error message (in title). Is there something I'm missing?

[lsorber]

Hi @antara-c, thanks for the report. Could you provide a minimal reproducible example please?

[antara-c]

Hi @lsorber , sure here's the bit that throws an error to me

forecaster = DartsForecaster(
model=conformal_predictor,
lags=5 * 24, # Add the last 5 days of the target to the prediction features
lags_future_covariates=[0], # Add the current timestamp's covariates to the prediction features
#categorical_future_covariates=X_categoricals, # Convert these covariates to pd.Categorical
)

Fit the forecaster

forecaster.fit(y1_train, future_covariates=X1_train)

Make a probabilistic forecast 5 days into the future by predicting a set of conformally calibrated

quantiles at each time step and drawing 500 samples from them

quantiles = (0.025, 0.05, 0.1, 0.25, 0.5, 0.75, 0.9, 0.95, 0.975)
forecast = forecaster.predict(
n=5 * 24, future_covariates=X1_test, num_samples=500, quantiles=quantiles
)

I'm not using categorical variables
This is the error

AttributeError Traceback (most recent call last)
Cell In[67], line 1
----> 1 forecaster = DartsForecaster(
2 model=conformal_predictor,
3 lags=5 * 24, # Add the last 5 days of the target to the prediction features
4 lags_future_covariates=[0], # Add the current timestamp's covariates to the prediction features
5 #categorical_future_covariates=X_categoricals, # Convert these covariates to pd.Categorical
6 )
8 # Fit the forecaster
9 forecaster.fit(y1_train, future_covariates=X1_train)

File ~\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\LocalCache\local-packages\Python311\site-packages\conformal_tights_darts_forecaster.py:104, in DartsForecaster.init(self, model, lags, lags_past_covariates, lags_future_covariates, output_chunk_length, output_chunk_shift, add_encoders, multi_models, use_static_covariates, categorical_past_covariates, categorical_future_covariates, categorical_static_covariates)
102 # Initialise _LikelihoodMixin.
103 self.likelihood = "quantile"
--> 104 self._model_container = self._get_model_container()
105 self._rng = check_random_state(model.random_state) # Generator for sampling.
106 # Initialise darts.models.RegressionModel.

AttributeError: 'DartsForecaster' object has no attribute '_get_model_container'

[lsorber]

Hi @antara-c, I'll need a bit more information to be able to help you with this:

1. Could you share the output of pip list in your environment?
2. Could you provide a full example that I can run locally to reproduce the issue? This can be a .py or .ipynb file.

[lsorber]

Hi @antara-c, we just released v0.4.0 which brings a number of improvements. Specifically, #26 may help address the issue you reported. If that solves it for you, please let me know so that I can close this issue, thank you! If not, please provide the information I requested above 👆.

[antara-c]

hi, I upgraded the version of conformal tights and still get the above error

pip list as you requested
Package Version

---

adagio 0.2.4
aiohttp 3.9.5
aiosignal 1.3.1
antlr4-python3-runtime 4.11.1
appdirs 1.4.4
asttokens 2.4.1
attrs 23.2.0
catboost 1.2.5
certifi 2024.2.2
charset-normalizer 3.3.2
cloudpickle 3.0.0
colorama 0.4.6
comm 0.2.2
conformal-tights 0.4.0
contourpy 1.2.1
coreforecast 0.0.8
cycler 0.12.1
Cython 3.0.10
debugpy 1.8.1
decorator 5.1.1
executing 2.0.1
filelock 3.13.4
fonttools 4.51.0
frozenlist 1.4.1
fs 2.4.16
fsspec 2024.3.1
fugue 0.8.7
fugue-sql-antlr 0.2.0
graphviz 0.20.3
h5py 3.11.0
holidays 0.47
idna 3.7
intel-openmp 2021.4.0
ipykernel 6.29.4
ipython 8.23.0
ipywidgets 8.1.2
jedi 0.19.1
Jinja2 3.1.3
joblib 1.4.0
jupyter_client 8.6.1
jupyter_core 5.7.2
jupyterlab_widgets 3.0.10
kiwisolver 1.4.5
lightning-utilities 0.11.2
llvmlite 0.42.0
MarkupSafe 2.1.5
matplotlib 3.8.4
matplotlib-inline 0.1.7
missingno 0.5.2
mkl 2021.4.0
mpmath 1.3.0
multidict 6.0.5
nest-asyncio 1.6.0
networkx 3.3
nfoursid 1.0.1
numba 0.59.1
numpy 1.26.4
packaging 24.0
pandas 2.2.2
parso 0.8.4
patsy 0.5.6
pillow 10.3.0
platformdirs 4.2.1
plotly 5.22.0
pmdarima 2.0.4
prompt-toolkit 3.0.43
protobuf 5.26.1
psutil 5.9.8
pure-eval 0.2.2
pyarrow 16.0.0
Pygments 2.17.2
pyod 1.1.3
pyparsing 3.1.2
python-dateutil 2.9.0.post0
pytorch-lightning 2.2.3
pytz 2024.1
pywin32 306
PyYAML 6.0.1
pyzmq 26.0.2
qpd 0.4.4
requests 2.31.0
roerich 0.6.0
ruptures 1.1.9
scikit-base 0.7.7
scikit-learn 1.4.2
scipy 1.13.0
seaborn 0.13.2
shap 0.45.0
six 1.16.0
sktime 0.28.0
slicer 0.0.7
sqlglot 23.11.2
stack-data 0.6.3
statsforecast 1.7.4
statsmodels 0.14.2
sympy 1.12
tbats 1.1.3
tbb 2021.12.0
tenacity 8.2.3
tensorboardX 2.6.2.2
threadpoolctl 3.4.0
torch 2.3.0
torchaudio 2.3.0
torchmetrics 1.3.2
torchvision 0.18.0
tornado 6.4
tqdm 4.66.2
traitlets 5.14.3
triad 0.9.6
typing_extensions 4.11.0
tzdata 2024.1
u8darts 0.29.0
urllib3 2.2.1
utilsforecast 0.1.7
wcwidth 0.2.13
widgetsnbextension 4.0.10
xarray 2024.3.0
xgboost 2.0.3
yarl 1.9.4

I will not be able to share the py file due to constraints. I am able to produce the first graph though (https://pypi.org/project/conformal-tights/) so I would understand if you need to close this due to the lack of the source code. Thanks

[lsorber]

All right, I'd like to help but it's fairly difficult to debug this without having an example that I can use to reproduce the issue locally.

If you want to proceed, I'd need to know the following:

1. Can you share how you created the conformal_predictor you passed to DartsForecaster?
2. Is isinstance(forecaster, darts.models.forecasting.regression_model._LikelihoodMixin) true?
3. Is hasattr(darts.models.forecasting.regression_model._LikelihoodMixin, "_get_model_container") true?
4. Is hasattr(forecaster, "_get_model_container") true?

The answers to (2)-(4) should be yes. If the answer is no to any of these, that will give us a clue as to where the issue is.