R/mhealth.convert.R
mhealth.convert.RdConvert dataframe to mhealth specification. The column order of converted dataframe would be required_cols, group_cols. Columns that are not specified in the input arguments will be dropped.
mhealth.convert(df, file_type, required_cols, rename_cols = NULL, group_cols = NULL, datetime_format = NULL, timezone = "UTC")
| file_type | mhealth specification file types, see `mhealth$filetype` for all supported types |
|---|---|
| required_cols | the required columns of file types that are in order. E.g. for annotation, timestamp, start and end time and annotation name columns are required in the order; for sensor, timestamp and necessary numeric columns are required; for feature, timestamp, start and end time and numeric or categoric features are required; for event, timestamp, start and end time and numeric or categoric info columns are required. |
| rename_cols | rename column names in results as provided string vector (except for the first timestamp column) |
| group_cols | the columns used for indexing or grouping. Will be converted to numeric or if not, character. Default is null, which means no group columns. |
| datetime_format | in string. Used to convert tiemstamp and start and stop time columns to date object |
| timezone | in location string. Used to convert timestamp and start and stop time columns to date object |