Question on Memory Consumption

Hello there! I want to determine the memory consumption when building a brotli.Compressor object utilizing the highest dictionary level (i.e. 11) and an empty dictionary (i.e. b"") for a float16 numpy array of 2k elements.

First things first, i'm approaching this via memory-profiler. An occuring problem is that, when i'm building the brotli.Compressor given the above attributes, i get an error that compressor obeject has no atribute process.

When i'm using compressed = brotli.compress(data,quality,mode,dict), the memory consumption seems to be significantly low.

Therefore:

1. Do you know an approximation of the memory used building the above described object?
2. How can i resolve the error compressor object has no attribute process?

Inizia verificando l’API brotli.Compressor e la configurazione di memory-profiler descritte nell’issue, quindi confrontale con la chiamata a brotli.compress usando l’input NumPy float16 indicato. Il lavoro è completo quando è stato stabilito se process è supportato e sono state documentate le modalità per misurare l’uso della memoria dell’oggetto richiesto o riprodurre la limitazione.