Changelog¶

0.4.0 - 2021-Feb-02¶

Interactive plot in strymread
Default map for strymmap set to Mapbox with an option for Googlemap
Now strymread function plt_ts can take a non-default column other than ‘Message’ for plotting as an argument
Bug fixes

Bug fixes in strymread
Strym tools
- graham_scan function to find convex hull and its perimeter for a cluster
- ellipse_fit function to fit an ellipse/circle to a cluster of points on phase-space

Documentation bug fixes
Plotting bug fixes and improvement
Strym tools
- ACD Tool added to measure stop-and-go distance. Animation generation for phasespace evolution supported
- Acceleration messages added to strymread to be dbc independent
- DBC now supports headlight message for highbeam, lowbeam etc.

Standalone functions from all classes moved to respective class definition as static method to reflect their usability
Sphinx version updated
Strym now packages DBC file. Currently suppported DBC file. Currently supported DBC file: Honda Pilot 2017, Toyota RAV4 2019 and 2020
Support for creating serverless sqlite3 database in local system for easy querying of driving information
class: strymread
- Bug fix: removes duplicate timestamped data from signals
- get_ts now has additional column BUS that provides number of each timestamped message
- Option to save RAW, UNDECODED Dataframe as sqlite3 via createdb parameter while object instantiation
- A helper message lead_distance to retrive lead distance from csv file
- Function name extract changed to export2mat that creates matlab .mat file of known, import signals
- state_space function to extract consolidated dataframe with a number of signals that represents state space of the vehicle, resampled to desire frequency. Optional argument todb, when True saves state space dataframe to a sqlite3 table STATE_SPACE
- differentiate method automatically falls back to AutoEncoder method when number of datapoints is less than 6.

class: strymread
- Fixed a scaling bug in prediction made in AE-based differentiation

modified for robust calling of meta and dashboard from snakemake files
class meta
- Takes an array of dbc files, and calls methods from the vin_parser package to confirm that the VIN of a CSV file shoudl correspond to a particular dbc
class strymread
- Added new functionality to do dictionary lookup of message/signal pairs when extracting a timeseries, rather than assuming all message/signal pairs correspond to Toyota naming conventions. New methods to touch when adding new message/signal pairs are prefixed with _dbc_

setup modified to package examples and README folder in install path
class dashboard
- works within the strym package to collect metadata files from within a folder and print interesting aspects of the collection
class meta
- Works within the strym package to extract metadata from drives that are recorded using libpanda, with optional corresponding dashcam video
class strymread
- Checks for monotonicity of the time in the recorded data
- Two new Flag Attributes added: burst and success. Burst flag tells if data was recorded in burst, success tells if reading of csv file was successful
class strymmap
- Conditionally import of bokeh and other widget-based libraries in Jupyter only to making strym work from terminal

class strymread
- A new function extract for extracting data in .MAT format to work with matlab
- Changes to msg_subset function, now returns an object of type strymread with modified dataframe
- Changes to plotting functions

class strymread
- Modified function to read radar traces.
- Bug fix for Function :code: frequency
- New functions:
  
  msg_subset: For deriving a subset of original messages satisfying given criteria. See commit #819f2d6
  
  time_subset: For getting time-slices of original messages satisfying given criteria
  
  acc_state: Get the ACC (Adaptive Cruise Control) state of driving
  
  accel_x: Get the longitudinal acceleration. See commit #819f2d6
Address issue #4, and issue #5.
Function ts_sync modified to inherit sample from one of the two input dataframes.
New function :code: time_index and :code: timeslices for improved analysis of timeseries data

class strymmap
- Replaced gmplot mapping with bokeh plot for map
- Now saves map with drive route as png file using selenium and chormium webdriver

Added a new class :code: strymmap
- Added basic funcionality to read and parse CSV file containing GPS data obtained from Grey Panda.
- Save Map HTML file from GPS Data to show drive route

A function to resample non-uniformly sampled timeseries to uniformly sampled timeseries data
A function to differentiate timeseries data based on spline derivative method
A function to denoise timeseries data based on moving average
A function to perform temporal-splitting of timeseries dataframe
A function to return centroid of a phase-space cluster
A function to calculate average distance of a phase-space cluster from its centroid
Plotting utility for temporal violin plot
Can retrieve a timeseries message by given message ID/signal ID or message name/signal name

class strymread
- Get the message count
- Functions to retrieve yaw, acceleration, steer torque, steer rate, steering angle, steering fraction, wheel speeds, longitudinal and laternal measurements from Radar traces
- Get datarate statistics from CAN data
- Plot trajectory of driving based on Kinematic model
timeseries-sync of two timeseries data of different and non-uniform sampling period
Off-the-shelf integration function for timeseries data
Function to analyze data rate throughput of a particular message.
Visualize data distributionb through violin plot

Added a new class strymread
- Added basic functionality to Parse CSV-formatted CAN data captured usin comma.ai Panda and Giraffe connector.
- Plot timeseries data by message name

Real-time capturing and visualization of CAN data using comma.ai Panda and Giraffe connector.