| Type: | Package |
| Title: | Analysis and Prediction of Bicycle Rental Amount |
| Version: | 0.1.1 |
| Maintainer: | Jiwon Min <miw5281@gmail.com> |
| Description: | Provides functions for analyzing citizens' bicycle usage pattern and predicting rental amount on specific conditions. Functions on this package interacts with data on 'tashudata' package, a 'drat' repository. 'tashudata' package contains rental/return history on public bicycle system('Tashu'), weather for 3 years and bicycle station information. To install this data package, see the instructions at https://github.com/zeee1/Tashu_Rpackage. top10_stations(), top10_paths() function visualizes image showing the most used top 10 stations and paths. daily_bike_rental() and monthly_bike_rental() shows daily, monthly amount of bicycle rental. create_train_dataset(), create_test_dataset() is data processing function for prediction. Bicycle rental history from 2013 to 2014 is used to create training dataset and that on 2015 is for test dataset. Users can make random-forest prediction model by using create_train_model() and predict amount of bicycle rental in 2015 by using predict_bike_rental(). |
| License: | GPL-2 | GPL-3 [expanded from: GPL (≥ 2)] |
| Encoding: | UTF-8 |
| LazyData: | true |
| Imports: | ggplot2, lubridate, dplyr, randomForest, plyr, reshape2, RColorBrewer, drat |
| Suggests: | knitr, rmarkdown, tashudata |
| Additional_repositories: | https://zeee1.github.io/drat |
| VignetteBuilder: | knitr |
| RoxygenNote: | 7.1.1 |
| Depends: | R (≥ 3.5.0) |
| NeedsCompilation: | no |
| Packaged: | 2021-01-13 09:03:15 UTC; miw52 |
| Author: | Jiwon Min [aut, cre] |
| Repository: | CRAN |
| Date/Publication: | 2021-01-13 09:50:02 UTC |
Create training dataset on specific station for prediction
Description
A function to create training dataset on 'station_number' bicycle station by preprocessing bicycle rental history and weather data from 2013 to 2014.
Usage
create_test_dataset(station_number)
Arguments
station_number |
number that means the number of each station.(1 ~ 144) |
Value
a dataset containing feature and rental count data on 'station_number' station, 2013 ~ 2014
Examples
## Not run: test_dataset <- create_test_dataset(1)
Create training dataset on specific station for prediction
Description
A function to create training dataset on 'station_number' bicycle station by preprocessing bicycle rental history and weather data from 2013 to 2014.
Usage
create_train_dataset(station_number)
Arguments
station_number |
number that means the number of each station.(1 ~ 144) |
Value
a dataset containing feature and rental count data on 'station_number' station, 2013 ~ 2014
Examples
## Not run: train_dataset <- create_train_dataset(1)
Create random-forest training model for bicycle rental prediction.
Description
Create random-forest training model for bicycle rental prediction.
Usage
create_train_model(train_dataset)
Arguments
train_dataset |
Training dataset created by create_train_dataset() |
Value
random forest training model
Examples
## Not run: train_dataset <- create_train_dataset(3)
rf_model <- create_train_model(train_dataset)
## End(Not run)
Visualize amount of bicycle rental at each day of week.
Description
A function analyzing bicycle rental pattern on each day of week and visualizing analyzed result.
Usage
daily_bicycle_rental()
Examples
## Not run: daily_bicycle_rental()
Extract feature columns from train/test dataset
Description
Extract feature columns from train/test dataset
Usage
extract_features(data)
Arguments
data |
data with feature columns and others |
Value
data containing only feature columns
Visualize the change of bicycle rental amount by temperature and each month.
Description
A function drawing a plot that shows change of temperature and bicycle rental ratio in each month.
Usage
monthly_bicycle_rental()
Examples
## Not run: monthly_bicycle_rental()
Predict hourly Demand of bicycle in 2015.
Description
predict hourly amount of bicycle rental in 2015 using random forest algorithm. Create prediction model using 'train_dataset' and forecast demand of bicycle rental according to the condition of 'test_dataset'
Usage
predict_bicycle_rental(rf_model, test_dataset)
Arguments
rf_model |
random forest prediction model create by create_train_model() |
test_dataset |
testing dataset |
Value
test_dataset with predictive result.
Examples
## Not run: train_dataset <- create_train_dataset(3)
test_dataset <- create_test_dataset(3)
rf_model <- create_train_model(train_dataset)
test_dataset <- predict_bicycle_rental(rf_model, test_dataset)
## End(Not run)
Visualize Top 10 Pathes that were most used from 2013 to 2015.
Description
Visualize Top 10 Pathes that were most used from 2013 to 2015.
Usage
top10_paths()
Examples
## Not run: top10_paths()
Visualize top 10 stations that were most used from 2013 to 2015.
Description
Draw a plot that visualized most used top 10 stations on barchart.
Usage
top10_stations()
Value
Data frame that contains top 10 most used stations from 2013 to 2015
Examples
## Not run: top10_stations()