You could create a simple utility function that either converts the given value to a float if possible, or returns it as is:
def maybe_float(s):
try:
return float(s)
except (ValueError, TypeError):
return s
orig_list = ['data', '18', '17', '0']
the_list = [maybe_float(v) for v in orig_list]
And please don’t use names of builtin functions and types such as list
etc. as variable names.
Since your data actually has structure instead of being a truly mixed list of strings and numbers, it seems a 4-tuple of (str, float, float, float)
is more apt:
data_conv = (data_l[0], *(float(v) for v in data_l[1:]))
or in older Python versions
# You could also just convert each float separately since there are so few
data_conv = tuple([data_l[0]] + [float(v) for v in data_l[1:]])