This repository is used to initially generate datasets of ResNet for blind recognition of polar codes.

• Mode Selection: Allows for the selection between training and testing modes.
• SNR Range Configuration: Enables the specification of the Signal-to-Noise Ratio (SNR) range for dataset generation.

The script includes several key configuration options:

• reco_mode: Determines the recognition mode. Options include 'len', and 'prate'.
• trainortest: Specifies whether the dataset is for training ('train') or testing ('test').
• no_samples_total: The total number of samples to generate. This value adjusts based on the mode and conditions.
• snr_range: The range of SNR values for which the dataset is generated. Configurable to suit various testing and training needs.
• N: The base value for N, configurable based on the reconstruction mode.

To use this script, simply configure the parameters at the top of the script to suit your dataset generation needs. The script will then generate a dataset based on these parameters, suitable for use in simulations.