I would like to convert data of double type to single type before code generation in 2020A Simulink, without using Cast blocks on most functions.

I would like to do this without using Cast blocks on most functions.

[Scott McNeil]

And what is the best way to build a model from the beginning in a smaller data type like uint16 wihout using many Cast blocks and Converters?

Answer 1

[Fangjun Jiang]

See this example

Validate a Single-Precision Model

[Scott McNeil]

Very interesting. I will try that tomorrow. Thank you.

[Fangjun Jiang]

The code generated after running set_param(gcs, 'DataTypeOverride', 'Single') shows using real32_T, which is float (single).

[Scott McNeil]

Thank you Fangjun Jiang. This was very helpful to get our model into Single Floating point which was very helpful to the Software coding process. I have also learned that changing to Fixed point types requires much more involvement during the code generating process. Some of the tools shared in the article were very good to learn.