Converting to pandas datetime data does not match format

I have a netcdf dataset and I want to convert the time to pandas datetime format.
The entries look like this:

<xarray.DataArray 'time' ()>
array(19810101.47916667)
Coordinates:
    time     float64 1.981e+07
Attributes:
    axis:           T
    calendar:       proleptic_gregorian
    standard_name:  time
    units:          day as %Y%m%d.%f

When I try to convert the data like this

pd.to_datetime(19810101.47916667,format='%Y%m%d.%f')

I get the error "time data ‘19810101’ does not match format ‘%Y%m%d.%f’ (match)".
I fail to undertand what exactly is going wrong here. Especially since this is exactly the format mentioned in the netcdf data (as you can see above).
Can someone tell me what I am missing?

>Solution :

You simply need to make your date value a string as below:

pd.to_datetime("19810101.47916667",format='%Y%m%d.%f'