estndv 0.0.2

Learned sample-based estimator for number of distinct values.

Learned NDV estimator

Learned model to estimate number of distinct values (NDV) of a population using a small sample. The model approximates the maximum likelihood estimation of NDV, which is difficult to obtain analytically. See our VLDB 2022 paper Learning to be a Statistician: Learned Estimator for Number of Distinct Values for more details.

How to use

1. Install the package

   pip install estndv

2. Import and create an instance

   from estndv import ndvEstimator
   estimator = ndvEstimator()

4. Assume your sample is S=[1,1,1,3,5,5,12] and the population size is N=100000. You can estimate population ndv by:

   ndv = estimator.sample_predict(S=[1,1,1,3,5,5,12], N=100000)

5. If you have the sample profile e.g. f=[2,1,1], you can estimate population NDV by:

   ndv = estimator.profile_predict(f=[2,1,1], N=100000)

6. If you have multiple samples/profiles from multiple populations, you can estimate population NDV for all of them in a batch by method estimator.sample_predict_batch() or estimator.profile_predict_batch().

How to train the ndv estimator

You can directly use our packaged library for your datasets, as the pre-trained model in the packed library is agnostic to any workloads. However, if you want to train the model from scratch anyway, do the following:

1. Go to the model_training folder cd model_training

2. Install requirements

   pip install requirements.txt

3. Generate training data. (This uses a lot of memory.)

   python training_data_generation.py

4. Train model

   python model_training.py

5. Save trained pytorch model parameters to numpy, this generates a file model_paras.npy

   python torch2npy.py

6. Test with your model parameters by specifying a path to your model_paras.npy

   estimator = ndvEstimator(para_path=your path to model_paras.npy)