User Guide
Single-precision floating-point number type, compatible with C float.
float
'f'
numpy.single
numpy.float32: 32-bit-precision floating-point number type: sign bit, 8 bits exponent, 23 bits mantissa.