Caar

- Added parameter `include_first_and_last_days=False` to consecutive_days_of_observations()
- Renamed `daily_cycle_and_temp_obs_counts` function to `daily_cycle_sensor_and_geospatial_obs_counts`
- Name variable names in test files consistent by replacing `inside` to `sensor_obs` and `outside` to `geosptial_obs`, to be more general

- ID and start/end arguments now optional in timeseries module functions (default to selecting all records)
- Optional ID and start/end arguments are keyword arguments instead of positional arguments
- Cycling device IDs and sensor IDs are independent parameters in timeseries module
- More general data category names for named tuple records in dicts and functions (Sensors, Geospatial)
- Sensor and geospatial data in timeseries module can be vectors or, as in previous releases, scalars

Fixed function to make it general for all data types in support of future API

- Added cols= parameter to temps_arr_by_freq in timeseries module for selecting column(s) by label

- Index column detection fix

- Added new parameters to pickle and dict creation functions for specifying ID or cycle mode column headings
- Made int, float and datetime.datetime conversion from strings part of pickle/dict creation
- Added column metadata to pickled dicts and dicts, pairing them and records dict within overall dict
- Made DataFrames retain original column headings from text files across all index and data columns
- Added cols_to_ignore, which takes either string or integer arguments
- Added df_select_datetime_range function for slicing DataFrames
- Added cycle mode to MultiIndex in cycling records and DataFrames