Average Travel Speed

Across all modes the average trip length was 17.8020611 miles, the average trip duration was 0.4366111 hours, the average speed was 36.7363008 MPH. The tables on the tab below show the break downs by mode as well as the whole data set.

Tables

By Mode

mode mean_distance mean_duration mean_speed
BIKING 2.4081942 0.2145415 11.347528
CARPOOL 24.3892918 0.5083140 46.366814
COMMERCIAL 18.4458339 0.3966075 43.079876
ON_DEMAND_AUTO 14.9032310 0.3388794 42.115975
PRIVATE_AUTO 24.6417151 0.5076874 46.406134
PUBLIC_TRANSIT 7.6634842 0.7737123 9.401329
WALKING 0.5455021 0.1676981 3.289728

Dataset

Proximity and Speed

Proximity and speed appear to have a non-linear relationship, most likely a logarithmic relationships. Because of this the lines of fit use a General Additive Model instead of a General Linear Model. The whole data set is plotted below.

The relationship appears to vary significantly by mode with public transit, walking, carpooling, and biking have significantly more linear relationships.

Binning by Proximity

A kmeans clustering process was applied to group the data by proximity. The overall distribution of proximity is shown below by mode.

Clusters are shown below as well as summary tables

cluster min_dist max_dist
1 0.000000 5.262779
2 5.271627 11.965238
3 11.977529 19.408520
4 19.422817 27.925792
5 27.943365 36.941061
6 36.957738 47.400000
7 47.426714 61.213176
cluster mode share
1 BIKING 0.0683802
1 CARPOOL 0.0147059
1 COMMERCIAL 0.0654143
1 ON_DEMAND_AUTO 0.3145401
1 PRIVATE_AUTO 0.0045553
1 PUBLIC_TRANSIT 0.2620519
1 WALKING 0.0062453
2 BIKING 0.4307692
2 CARPOOL 0.0060782
2 COMMERCIAL 0.0250864
2 ON_DEMAND_AUTO 0.1563126
2 PRIVATE_AUTO 0.0023353
2 PUBLIC_TRANSIT 0.1982759
3 CARPOOL 0.0154069
3 COMMERCIAL 0.0229245
3 ON_DEMAND_AUTO 0.1743440
3 PRIVATE_AUTO 0.0146148
3 PUBLIC_TRANSIT 0.4240000
4 CARPOOL 0.0910708
4 COMMERCIAL 0.0707779
4 ON_DEMAND_AUTO 0.3153153
4 PRIVATE_AUTO 0.0625463
4 PUBLIC_TRANSIT 0.6956522
5 CARPOOL 0.1572867
5 COMMERCIAL 0.1371974
5 ON_DEMAND_AUTO 0.4095238
5 PRIVATE_AUTO 0.0915901
5 PUBLIC_TRANSIT 0.6250000
6 CARPOOL 0.2453596
6 COMMERCIAL 0.2124646
6 ON_DEMAND_AUTO 0.6800000
6 PRIVATE_AUTO 0.1307357
6 PUBLIC_TRANSIT 1.0000000
7 CARPOOL 0.2665726
7 COMMERCIAL 0.3217391
7 ON_DEMAND_AUTO 1.0000000
7 PRIVATE_AUTO 0.1640576

Since the kmeans clusters didn’t fit your requirements I added a 0 cluster which was trips under 2 miles. Which creates the following break down.

cluster min_dist max_dist
0 0.000000 1.994866
1 2.000311 5.262779
2 5.271627 11.965238
3 11.977529 19.408520
4 19.422817 27.925792
5 27.943365 36.941061
6 36.957738 47.400000
7 47.426714 61.213176
cluster mode share
0 BIKING 0.0840395
0 COMMERCIAL 0.2000000
0 ON_DEMAND_AUTO 1.0000000
0 PRIVATE_AUTO 0.0148515
0 PUBLIC_TRANSIT 1.0000000
0 WALKING 0.0061646
1 BIKING 0.0617738
1 CARPOOL 0.0147059
1 COMMERCIAL 0.0575284
1 ON_DEMAND_AUTO 0.3063063
1 PRIVATE_AUTO 0.0040835
1 PUBLIC_TRANSIT 0.2509410
1 WALKING 0.3170732
2 BIKING 0.4307692
2 CARPOOL 0.0060782
2 COMMERCIAL 0.0250864
2 ON_DEMAND_AUTO 0.1563126
2 PRIVATE_AUTO 0.0023353
2 PUBLIC_TRANSIT 0.1982759
3 CARPOOL 0.0154069
3 COMMERCIAL 0.0229245
3 ON_DEMAND_AUTO 0.1743440
3 PRIVATE_AUTO 0.0146148
3 PUBLIC_TRANSIT 0.4240000
4 CARPOOL 0.0910708
4 COMMERCIAL 0.0707779
4 ON_DEMAND_AUTO 0.3153153
4 PRIVATE_AUTO 0.0625463
4 PUBLIC_TRANSIT 0.6956522
5 CARPOOL 0.1572867
5 COMMERCIAL 0.1371974
5 ON_DEMAND_AUTO 0.4095238
5 PRIVATE_AUTO 0.0915901
5 PUBLIC_TRANSIT 0.6250000
6 CARPOOL 0.2453596
6 COMMERCIAL 0.2124646
6 ON_DEMAND_AUTO 0.6800000
6 PRIVATE_AUTO 0.1307357
6 PUBLIC_TRANSIT 1.0000000
7 CARPOOL 0.2665726
7 COMMERCIAL 0.3217391
7 ON_DEMAND_AUTO 1.0000000
7 PRIVATE_AUTO 0.1640576

Mode Share Maps

The following tabs show the polygons by mode share, each tab presenting a different mode share.

The coverage appears to be better in Broward county, which makes sense as that was the project this was intended for. My guess is that is trips originating in Broward County. I’ll need to dig into how it was collected to confirm that there is missing data elsewhere.

Maps

WALKING
BIKING
CARPOOL
COMMERCIAL
ON_DEMAND_AUTO
PRIVATE_AUTO
PUBLIC_TRANSIT

Cluster Maps

Because each block group can be apart of multiple clusters due to multiple modes the clusters are presented by travel modes below.

Maps

WALKING
BIKING
CARPOOL
COMMERCIAL
ON_DEMAND_AUTO
PRIVATE_AUTO
PUBLIC_TRANSIT