niimpy.preprocessing.tracker module¶
- niimpy.preprocessing.tracker.extract_features_tracker(df, features=None)[source]¶
This function computes and organizes the selected features for tracker data recorded using Polar Ignite.
The complete list of features that can be calculated are: tracker_daily_step_distribution
- Parameters
- df: pandas.DataFrame
Input data frame
- features: dict, optional
Dictionary keys contain the names of the features to compute. The value of the keys is the list of parameters that will be passed to the function. If none is given, all features will be computed.
- Returns
- result: dataframe
Resulting dataframe
- niimpy.preprocessing.tracker.step_summary(df, value_col='values', user_id=None, start_date=None, end_date=None)[source]¶
Return the summary of step count in a time range. The summary includes the following information of step count per day: mean, standard deviation, min, max
- Parameters
- dfPandas Dataframe
Dataframe containing the hourly step count of an individual. The dataframe must be date time index.
- value_col: str.
Column contains step values. Default value is “values”.
- user_id: list. Optional
List of user id. If none given, returns summary for all users.
- start_date: string. Optional
Start date of time segment used for computing the summary. If not given, acquire summary for the whole time range.
- end_date: string. Optional
End date of time segment used for computing the summary. If not given, acquire summary for the whole time range.
- Returns
- summary_df: pandas DataFrame
A dataframe containing user id and associated step summary.
- niimpy.preprocessing.tracker.tracker_daily_step_distribution(steps_df)[source]¶
Return distribution of steps within each day. Assuming the step count is recorded at hourly resolution, this function will compute the contribution of each hourly step count into the daily count (percentage wise).
- Parameters
- steps_dfPandas Dataframe
Dataframe containing the hourly step count of an individual.
- Returns
- df: pandas DataFrame
A dataframe containing the distribution of step count per day at hourly resolution.