Transportation Services commissioned the Toronto Cycling App by Brisk Synergies in 2014. The App allows cyclists to record their cycling routes and provide this data to the City. This data was used to inform the development of the new Cycling Plan and also assists in the ongoing monitoring of cycling patterns over time as cycling infrastructure is improved and expanded.
The following data dumps are available from Brisk's analytics portal.
trip_surveys.csv
user_surveys.csv
Additionally, for the TrafficJam Hackathon were provided:
- daily od-matrices (TTS 2006 TAZ)
- a shapefile of volume by links (TCL)
"2017" data has 1867 rows There is overlap in trips between 2014 & 2015 files
trip_id: bigint
app_user_id:
started_at: timestamp in UTC, 2017 file from 2016-01-25 to 2017-02-02
purpose: trip purpose, see below table
notes: Mostly empty, otherwise trip purpose for "other", some are in json format like: {"comments":"test"}
| Purpose | Trips |
|---|---|
| Autre | 0.05% |
| Commute | 62.22% |
| Errand | 2.41% |
| Exercise | 1.18% |
| Home | 5.14% |
| Leisure | 2.41% |
| Loisirs | 0.05% |
| Other | 9.43% |
| School | 3.43% |
| Shopping | 4.82% |
| Social | 5.25% |
| Workrelated | 0.16% |
| Work-related | 3.43% |
7934 rows, spanning ids from 1 to 8094. Previous years' portals have a subset of the most recent file. Some of the values for the variables don't seem to match the answer key in the json below.
| column | type | notes |
|---|---|---|
| app_user_id | int | |
| winter | [0-2] | see below |
| rider_history | [0-4] | not defined |
| workZIP | varchar(7) | Postal Code |
| income | [0-6] | see below |
| cyclingFreq | 0 or null | see below |
| age | [0-7] | see below |
| cycling_level | [0-3] | see below |
| gender | [0-3] | see below |
| rider_type | 0 or null | see below |
| schoolZIP | varchar(7) | Postal Code |
| homeZIP | varchar(7) | Postal Code |
| cyclingExperience | [0-4] | see below |
| preference_key_userpref | boolean | TRUE or null |
The preference_key_userpref column starts at app_user 6821, only three users have the "True" value. The number of commas in the csv wasn't consistent, so use python/fix_user_surveys.py to correct this.
The dictionary for most of the answers are in the json below. See sql/insert_json_data.sql for a way of automating creation of a normalized table structure for each variable in the json.
"cyclingfreq": {
"0": "Less than once a month",
"1": "Several times a month",
"2": "Several times per week",
"3": "Daily"
},
"age": {
"1": "Less than 18",
"2": "1824",
"3": "2534",
"4": "3544",
"5": "4554",
"6": "5564",
"7": "65"
},
"gender": {
"1": "Female",
"2": "Male",
"3": "Other",
"4": "Not Specified"
},
"winter": {
"1": "No",
"2": "Yes"
},
"income": {
"1": "Less than 20,000",
"2": "20,000 to 39,999",
"3": "40,000 to 59,999",
"4": "60,000 to 74,999",
"5": "75,000 to 99,999",
"6": "100,000 or greater"
},
"rider_type": {
"1": "Strong & fearless",
"2": "Enthused & confident",
"3": "Comfortable, but cautious",
"4": "Interested, but concerned"
},
"cyclingExperience": {
"1": "Since childhood",
"2": "Several years",
"3": "One year or less",
"4": "Just trying it out just started"
},
"cycling_level": {
"1": "Not comfortable riding with traffic",
"2": "Only comfortable with dedicated facilities",
"3": "Comfortable riding with traffic"
}Download portal requires you to enter a bounding box and start and end date for trip filtering. Can only download 1000 trips at a time. We got a month of data for TrafficJam which was ~1GB.
Fields (all values in the csv are quoted)
| column | type | notes |
|---|---|---|
| coord_id | bigint | serial for GPS ping |
| trip_id | bigint | trip uid |
| recorded_at | timestamp | timezone unclear (probably UTC) |
| longitude | numeric | 6 decimal |
| latitude | numeric | 6 decimal |
| altitude | numeric | appears to be meters |
| speed | numeric | Avg: 4.81 so probably km/h? |
| hort_accuracy | numeric | probably meters |
| vert_accuracy | numeric | probably meters |