Encode all unable to use 'f32' or 'f16' formats

[HeatPhoenix]

Hi,

In line 148 of the file encode.py, you seem to have made a small error in the sense that you initialize a default ValuesEncoder. This means that if you are trying to do encoding in a different format (say 'f32') this will be superceded by the default. In other words, this means that in the current pip implementation of gorillacompression it is impossible to do anything but 'f64' compression unless you use encode_next manually.

Simple fix would be to initialize that value encoder in that function using the values from "self.". (Alternatively, I can make a pull request).

Regards,
Zack

[ghilesmeddour]

Hi @HeatPhoenix,

I understand what you mean, encode_all encodes f64 by default and does not allow you to choose a different encoding. Thanks for bringing this up.

We can't use self because encode_all is static, but we can add float_format parameter to encode_all.

[ghilesmeddour]

@HeatPhoenix I have just published a new release with the fix