roboscout
Requirements:
- Flask https://flask.pocoo.org/
- termcolor https://pypi.python.org/pypi/termcolor
- and of course Python 2.7
How to use it:
- go to the
gatheringfolder and runpython main.pyto start the gathering server. - any computer that connects to the IP address of the machine running that server on port 5000 will see the data gathering screen below (if the ip is
10.0.1.25, then type10.0.1.25:5000into your browser) - use the gathering page to submit data on each team (you can safely ignore final score for now)
- once you have gathered data and want to analyze it, copy
gathering/data/data.dbintoanalysis/data/data.dbusing whatever method you see fit: airsync, rsync, just manual copying, symlinking, etc. - then in the analysis folder, there are several scripts (see the README there for info about these scripts)
- there are a few config values that can be changed:
COLORS = True ## enable colored printing of output; disable if you see random characters
AUTOCOPY = False ## automatically copies the database from the gathering folder when an analysis program is runThese are located at the top of the main config file in the root directory.
Data acquisition:
Team analysis:
Some planning I did a while ago about this app:
- gets data like number of goal crossings, crossing speed, boulder stats, challenge and capture of tower, auton stats, and final game score
- run on a raspberry pi
- backpack in the stands with several ethernets coming out of it:
- contains raspberry pi, router, power source, and some communication method back to pit area
- maybe we can run an ac power cord, if not, big multirotor batteries with a switching power supply
- need to find an 8-port router for linking the laptops together
- so there are 6 laptops linked to this hub, all accessing the raspberry pi’s webserver
- the webserver serves a page to enter various data about the specific bot
- each scouter will fill out one form per match, with 6 computers being able to fill out the data on the whole alliance (including our own bot when we’re there)
- the raspberry pi will then periodically sync the database back to the pit area computer (a more powerful one), which will do a bunch of analysis on it
- alternately, there is a 7th laptop at the scouting camp that does all the analysis, and the drive team confers with it before the match (no long-distance syncing needed)
- Match-based data for planning:
- {opposing, partner} alliance strengths and weaknesses: outer works defense crossing ablilties, defense, offense low and high
- suggested roles for us and our teammates’ bots (defense, offense low, offense high, supply), best opposing strategy
- suggested outerworks defense selections
- suggested robot placement for auton
- match history overview for each bot
- estimated outcome
- Robot-based data:
- strengths and weaknesses, best and worst roles
- best opposing strategy
- past match preformance
- Finals selection suggestion data:
- (this could take a while to compute, we’d run this analysis between quals and final selections, need to figure out how fast it can be done)
- top teams in general
- top teams that pair with our skills (or the skills of any given bot)
- top alliances (selected for most skills)
- what the computer will do is run every single possible combination of teams in alliances of 3, and compare these results numerically
- easy way to view and edit historic data in case a scouter makes a mistake
- possibility of sharing this data with other teams?
- of course we’d share upcoming match analysis with our alliance mates
- maybe share finalist selection suggestions?
- pit scouting is better suited for another app or paper
- use pit scouting and practice field scouting until we have enough of this data to go off of
also a full(ish) printout (with actual data), lacking colors.
$ python match.py 539,1719,1123 5587,623,1418
BEGIN DATA PRINTOUT
TEAM 539
==== Defense ratings: ====
lowbar 12 (100.0%) ##################################################
portcullis 5 (41.6%) ####################
cheval 3 (20.0%) ##########
sallyport 2 (11.6%) #####
roughterrain 0 (0.0%)
moat 0 (0.0%)
drawbridge 0 (0.0%)
ramparts 0 (0.0%)
rockwall 0 (0.0%)
==== Best choices for defense: ====
Class D: rockwall at 0.0%
Class B: ramparts at 0.0%
Class C: drawbridge at 0.0%
Class A: cheval at 20.0%
==== Offense stats: ====
---- Low goal: ----
11 low goals made
average of 1.22222222222 per match
0 in auton
---- High goal: ----
0 high goals made, 0 missed (0 total)
average of 0.0 per match
successful 0.0% of the time
0 in auton
==== Auton stats: ====
0 low goals, 0 high gloals
started with a boulder 5 times (55.5555555556% of matches)
reached 6 times (66.6666666667% of matches)
crossed 3 times (33.3333333333% of matches)
==== Tower stats: ====
challenged tower 2 times (22.2222222222% of matches)
climbed tower 0 times (0.0% of matches)
partially climbed tower 0 times (0.0% of matches)
succeeded climbing tower 0% of attempts
TEAM 1719
==== Defense ratings: ====
rockwall 11 (100.0%) ##################################################
sallyport 6 (81.8%) ########################################
cheval 6 (54.5%) ###########################
lowbar 5 (45.4%) ######################
ramparts 5 (40.9%) ####################
roughterrain 3 (27.2%) #############
moat 2 (18.1%) #########
portcullis 0 (0.0%)
drawbridge 0 (0.0%)
==== Best choices for defense: ====
Class C: drawbridge at 0.0%
Class A: portcullis at 0.0%
Class B: moat at 18.1%
Class D: roughterrain at 27.2%
==== Offense stats: ====
---- Low goal: ----
0 low goals made
average of 0.0 per match
0 in auton
---- High goal: ----
0 high goals made, 0 missed (0 total)
average of 0.0 per match
successful 0.0% of the time
0 in auton
==== Auton stats: ====
0 low goals, 0 high gloals
started with a boulder 1 times (12.5% of matches)
reached 1 times (12.5% of matches)
crossed 0 times (0.0% of matches)
==== Tower stats: ====
challenged tower 1 times (12.5% of matches)
climbed tower 0 times (0.0% of matches)
partially climbed tower 0 times (0.0% of matches)
succeeded climbing tower 0% of attempts
TEAM 1123
==== Defense ratings: ====
lowbar 9 (100.0%) ##################################################
moat 8 (90.4%) #############################################
roughterrain 7 (66.6%) #################################
rockwall 6 (57.1%) ############################
portcullis 2 (28.5%) ##############
sallyport 1 (14.2%) #######
ramparts 1 (4.7%) ##
drawbridge 0 (0.0%)
cheval 0 (0.0%)
==== Best choices for defense: ====
Class A: cheval at 0.0%
Class C: drawbridge at 0.0%
Class B: ramparts at 4.7%
Class D: rockwall at 57.1%
==== Offense stats: ====
---- Low goal: ----
0 low goals made
average of 0.0 per match
0 in auton
---- High goal: ----
0 high goals made, 0 missed (0 total)
average of 0.0 per match
successful 0.0% of the time
0 in auton
==== Auton stats: ====
0 low goals, 0 high gloals
started with a boulder 1 times (12.5% of matches)
reached 2 times (25.0% of matches)
crossed 2 times (25.0% of matches)
==== Tower stats: ====
challenged tower 2 times (25.0% of matches)
climbed tower 0 times (0.0% of matches)
partially climbed tower 0 times (0.0% of matches)
succeeded climbing tower 0% of attempts
TEAM 5587
==== Defense ratings: ====
rockwall 9 (100.0%) ##################################################
portcullis 6 (66.6%) #################################
drawbridge 4 (44.4%) ######################
roughterrain 3 (33.3%) ################
lowbar 1 (3.7%) #
moat 0 (0.0%)
cheval 0 (0.0%)
ramparts 0 (0.0%)
sallyport 0 (0.0%)
==== Best choices for defense: ====
Class C: sallyport at 0.0%
Class B: ramparts at 0.0%
Class A: cheval at 0.0%
Class D: roughterrain at 33.3%
==== Offense stats: ====
---- Low goal: ----
0 low goals made
average of 0.0 per match
0 in auton
---- High goal: ----
0 high goals made, 0 missed (0 total)
average of 0.0 per match
successful 0.0% of the time
0 in auton
==== Auton stats: ====
0 low goals, 0 high gloals
started with a boulder 3 times (37.5% of matches)
reached 5 times (62.5% of matches)
crossed 4 times (50.0% of matches)
==== Tower stats: ====
challenged tower 1 times (12.5% of matches)
climbed tower 0 times (0.0% of matches)
partially climbed tower 0 times (0.0% of matches)
succeeded climbing tower 0% of attempts
TEAM 623
==== Defense ratings: ====
rockwall 18 (100.0%) ##################################################
portcullis 3 (25.7%) ############
moat 4 (22.8%) ###########
drawbridge 2 (17.1%) ########
lowbar 3 (11.4%) #####
ramparts 1 (2.8%) #
sallyport 1 (1.4%)
roughterrain 0 (0.0%)
cheval 0 (0.0%)
==== Best choices for defense: ====
Class A: cheval at 0.0%
Class D: roughterrain at 0.0%
Class C: sallyport at 1.4%
Class B: ramparts at 2.8%
==== Offense stats: ====
---- Low goal: ----
2 low goals made
average of 0.285714285714 per match
0 in auton
---- High goal: ----
0 high goals made, 0 missed (0 total)
average of 0.0 per match
successful 0.0% of the time
0 in auton
==== Auton stats: ====
0 low goals, 0 high gloals
started with a boulder 4 times (57.1428571429% of matches)
reached 0 times (0.0% of matches)
crossed 0 times (0.0% of matches)
==== Tower stats: ====
challenged tower 6 times (85.7142857143% of matches)
climbed tower 2 times (28.5714285714% of matches)
partially climbed tower 0 times (0.0% of matches)
succeeded climbing tower 100.0% of attempts
TEAM 1418
==== Defense ratings: ====
lowbar 13 (100.0%) ##################################################
portcullis 9 (62.8%) ###############################
ramparts 4 (31.4%) ###############
cheval 5 (25.7%) ############
sallyport 7 (17.1%) ########
roughterrain 1 (14.2%) #######
rockwall 1 (5.7%) ##
moat 0 (0.0%)
drawbridge 0 (0.0%)
==== Best choices for defense: ====
Class C: drawbridge at 0.0%
Class B: moat at 0.0%
Class D: rockwall at 5.7%
Class A: cheval at 25.7%
==== Offense stats: ====
---- Low goal: ----
11 low goals made
average of 1.0 per match
4 in auton
---- High goal: ----
6 high goals made, 2 missed (8 total)
average of 0.857142857143 per match
successful 75.0% of the time
0 in auton
==== Auton stats: ====
4 low goals, 0 high gloals
started with a boulder 4 times (57.1428571429% of matches)
reached 4 times (57.1428571429% of matches)
crossed 5 times (71.4285714286% of matches)
==== Tower stats: ====
challenged tower 1 times (14.2857142857% of matches)
climbed tower 1 times (14.2857142857% of matches)
partially climbed tower 0 times (0.0% of matches)
succeeded climbing tower 100.0% of attempts
ALLIANCE 539,1719,1123
==== Defense ratings: ====
lowbar 26 (100.0%) ##################################################
rockwall 17 (55.7%) ###########################
sallyport 9 (40.1%) ####################
moat 10 (37.7%) ##################
roughterrain 10 (32.7%) ################
portcullis 7 (30.3%) ###############
cheval 9 (29.5%) ##############
ramparts 6 (16.3%) ########
drawbridge 0 (0.0%)
==== Best choices for defense: ====
Class C: drawbridge at 0.0%
Class B: ramparts at 16.3%
Class A: cheval at 29.5%
Class D: roughterrain at 32.7%
ALLIANCE 5587,623,1418
==== Defense ratings: ====
rockwall 28 (100.0%) ##################################################
portcullis 18 (76.5%) ######################################
lowbar 17 (62.5%) ###############################
drawbridge 6 (28.1%) ##############
roughterrain 4 (21.8%) ##########
ramparts 5 (18.7%) #########
cheval 5 (14.0%) #######
moat 4 (12.5%) ######
sallyport 8 (10.1%) #####
==== Best choices for defense: ====
Class C: sallyport at 10.1%
Class B: moat at 12.5%
Class A: cheval at 14.0%
Class D: roughterrain at 21.8%
MATCH 539,1719,1123 vs 5587,623,1418
.........
END DATA PRINTOUT