thereallinusrg / cma-week3 Goto Github PK

View Code? Open in Web Editor NEW

0.0 0.0 0.0 657 KB

Exercise 3 of the course "Computational Movement Analysis"

HTML 100.00%

cma-week3's People

Contributors

Watchers

cma-week3's Issues

Frechet distance error

cma-week3/ex3_md.Rmd

Line 202 in aa1e24e

```

It seems there is an issue with that
see dionisanderegg/cma-week3#1

Task 5 _ plot not complete

cma-week3/ex3_md.Rmd

Lines 141 to 148 in aa1e24e

  
 pedestrian%>% 

  ggplot(aes(x= E, y= N)) + 

  geom_path(aes(colour=TrajID)) + 

  geom_point(aes(colour=TrajID)) +  

  facet_wrap(~TrajID, nrow = 2) + 

  coord_equal(1) +  

  theme_set(theme_minimal()) +

You correctly created the six different plots based on every of the available trajectories. In order to come up with a plot which resembles even more the one we have in the gitbook (https://computationalmovementanalysis.github.io/W03_04_tasks_and_inputs.html), you have to add also once more the dataset (pedestrians.csv), without including this time the TrajID info. Below I attach how I did it:

ggplot(pedestrians, aes(E,N)) +                                                                                   
  geom_point(data = dplyr::select(pedestrians, -TrajID),alpha = 0.1) +  # here I am adding the dataset once 
                                                                        # more, without the TrajID column                                          
  geom_point(aes(color = as.factor(TrajID)), size = 2) +                                                          
  geom_path(aes(color = as.factor(TrajID))) +                                                                     
  facet_wrap(~TrajID,labeller = label_both) +                                                                    
  coord_equal() +                                                                                                 
  theme_minimal() +                                                                                               
  labs(title = "Visual comparison of the 6 trajectories", subtitle = "Each subplot highlights a trajectory") +
  theme(legend.position = "none")

Task 4 _ segments < 5

cma-week3/ex3_md.Rmd

Lines 118 to 121 in aa1e24e

 caro5plus <- caro %>% 

  group_by(segment_id) %>% 

  mutate(n = length(TierID)) %>% 

  filter(n >= 5)

This is an accepted way to go, since there is the assumption that everything is sampled in a 1min interval.
You could also filter out the segments with duration < 5 mins, by cmputing the actual duration and then performing the filtering based on the newly created column. You could use the following expression to do it:

mutate(duration = as.integer(difftime(max(DatetimeUTC),min(DatetimeUTC),"mins")))

Something like this:

caro60_moves %>%                                                                   
  group_by(segment_id) %>%                                                              
  mutate(duration = as.integer(difftime(max(DatetimeUTC),min(DatetimeUTC),"mins"))) %>%
  filter(duration > 5) %>%                                                               
  ggplot(aes(E, N, color = segment_id)) +                                                
  geom_point() +                                                                         
  geom_path() +                                                                         
  coord_equal() +                                                                        
  theme(legend.position = "none") +                                                    
  labs(subtitle = "Long segments (removed segements <5 minutes)")

caro60.csv imported???

https://github.com/thereallinusrg/cma-week3/blob/aa1e24e623a66450956695e8f1fb04bfa8c846dbC:/Users/bako/Desktop/kurse/FS21/week3_groupBako/MarjanTababaka/cma3/Exercise3.R#L8-L8

I cannot see where and how you imported your caro60.csv file. I was expecting something like the following:

caro60 <- readr::read_delim("pathToYourFile",",")

reduce LCSS computational time

cma-week3/ex3_md.Rmd

Lines 204 to 206 in aa1e24e

 
 LCSS takes super long. Therefore i save the environment in case I need it again. 

 ```{r Save the Environment}

chek pampicloud/cma-week3#4 and our solutions (https://computationalmovementanalysis.github.io/W03_05_solutions.html) to see how we reduced computation time on LCSS

thereallinusrg / cma-week3 Goto Github PK

cma-week3's People

Contributors

Watchers

cma-week3's Issues

Frechet distance error

Task 5 _ plot not complete

Task 4 _ segments < 5

caro60.csv imported???

reduce LCSS computational time

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent


	pedestrian%>%
	ggplot(aes(x= E, y= N)) +
	geom_path(aes(colour=TrajID)) +
	geom_point(aes(colour=TrajID)) +
	facet_wrap(~TrajID, nrow = 2) +
	coord_equal(1) +
	theme_set(theme_minimal()) +

	caro5plus <- caro %>%
	group_by(segment_id) %>%
	mutate(n = length(TierID)) %>%
	filter(n >= 5)


	LCSS takes super long. Therefore i save the environment in case I need it again.
	```{r Save the Environment}