PerformanceWarning while saving to h5

[a-zmz]

while caching behaviour.Behaviour.get_positional_rate output, we get:

PerformanceWarning:
your performance may suffer as PyTables will pickle object types that it cannot
map directly to c-types [inferred_type->integer,key->axis0_level0] [items->None]

Both dfs has consistent data type float64, and index is int64. Tried to use pd.HDFStore to save data, but for big data it needs something specific.
So currently not sure why, and the only downside now is that it does not compress as much.