Challenge #76: Strava Export Parse and Report

Strava is a popular app that serves a social network for all athletes. In particular, the app is popular with runner and cyclists. However, sometimes we want to see the data differently from what is served to us. Below is a GPX file containing a mountain bike ride I completed a few weeks ago. The challenge is to parse it and create report snippets like these:

A speed and elevation dual-axis chart

A map of the route

Feel free to try anything else on top of these! I will not be providing a starting workflow since the results are pictured above, and connecting to the .gpx file is part of the challenge!

White_ranch.txt

Reporting

Parse

Spatial Analysis

Advanced

Data Analysis

Intermediate

Data Preparation

Spatial

Accepted answers

All comments

NicoleJohnson

My solution!

Spoiler

Finally found another opportunity to use the Make Columns tool!! Put the outputs into a pretty little report, motivational logo included. Makes me want to go ride a bike.

challenge_76_NicoleJohnson.yxmd

nick_ceneviva

Solution below

Spoiler

challenge_76_solution_Ceneviva.yxmd

gnans19

My Solution

Spoiler

challenge_76_solution_Gnans.yxmd

JSaumya

My solution. Lengthy but gives the output!

Challenge_76_solution.yxmd

pvara

Hi @JoeM how did you convert the gpx file to text? I am a avid strava user and wanted to convert all of my gpx files.

Thank you

JoeM

@pvara - when I downloaded it, it simply saved it with a .txt extension since .gpx is not an accepted Community file type. You can keep it as .gpx for Alteryx and process it as if it was a text file.

EmanueleE

It's my first attempt to the weeekly challenge. Hope to do it in the right way.

Challenge76-v1.yxmd

Garrett

The distance between lat/long observations divided by the time elapsed results in some pretty phenomenal speeds!

Or perhaps Joe can really climb Belcher Hill at 60mph? What's your secret Joe? Newfangled e-bike? Secret high speed chair lift?

strava_glodewyck.yxmd

Joe_Mako

My first draft is attached (named minimal), and upon looking at the data, I learned that in the Charting tool, the Chart Type of Line does not allow us to set an X-Axis, so what it uses is the record sequence. This means if we want to use Time as the X-Axis, each record should represent the same interval if we want an accurate representation of the data. In the data provided one record only represents one second for about half the records, the visualization in the thread question does not take these factors into account. Without taking them into account, the resulting chart is distorted.

Here is an example of the difference between Raw and Densified (note that with the Densified version we can see the moments when there was no movement, speed stays at 0 for a bit of time, and elevation is flat):

raw vs densified.png

Attached is my complex route with a way to densify the data with respect to each second of time, and then interpolate the coordinates and elevation. An additional chart with distance as the X-Axis was also created.

Side note, instead of using a Multi-Row Formula tool twice, one option is we can use a Join Multiple on record position keeping only records that join from all inputs, with one data stream skipping the first record. I am not sure if it is more efficient than multiple Multi-Row Formula tools.

Spoiler

c76 complex Joe_Mako.yxmd

c76 minimal Joe_Mako.yxmd

Laurap1228

Got to have some fun with the Reporting Tools on this one!

Spoiler

Challenge_76_LP.pdf

challenge_76_LP.yxmd

estherb47

My solution. I also filled in the time gaps for consistency, since the tools don't let you plot time on the horizontal axis.

Fun challenge!

Challenge_76_EHBSolution.yxmd

jamielaird

Here's my solution.

Spoiler

challenge_76-JL.yxmd

patrick_digan

Spoiler

Challenge_76_Start_file.yxmd

LordNeilLord

I forgot how bad the report tools were! Bring on plot.ly integration!

Spoiler

challenge_76_LNL.yxmd

SeanAdams

So - very similar solution to most of the solutions already provided

@JoeM - thank you for the XML exercise - the more XML and JSON and API work we can pack in, the better.

Spoiler

The bad:
- I didn't spot the date regularity issues that @Joe_Mako and @estherb47 used. Nice Catch!
- Also realise that I'm mixing units - elevation in meters but speed in MPH - @nick_ceneviva

the good:
But I did get a good excuse to play with the visual layout tool, and native XML importing using the input tool.
Also put the report together with some summary info on the top

The output:

Challenge76_SeanSolution.yxmd

RideReport.pdf

Natasha

As always, the most time was spent on making the charting tool to do what I want, which is frustrating especially after using Tableau on an every day basis.

Spoiler

I am not familiar with GPX files, but the first hit in Google told me that GPX, or GPS Exchange Format, is an XML schema designed as a common GPS data format for software applications.

I input provided txt file as XML file which saved me a lot of parsing. A little bit of metadata from the first rows was lost though it was not essential for this exercise.

Screen Shot 2017-11-12 at 15.22.40.png

challenge_76_Natasha.yxmd

LandonG

Solution attached.

challenge_76_start_file.yxzp

philipmannering

The reporting tools continue to confound, perplex and bamboozle me. Solution attached.

Spoiler

challenge_76 - Strava Parse.yxmd

ggruccio

Used tableau with a mapbox background to map the route. Created a dashboard with the elevation and map. Zipped file with workflow and tableau file below.