This repository has been archived by the owner on Jun 30, 2023. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 4
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Merge pull request #228 from jds485/227-txt-attrs-segs
add txt files with attribute and reach names
- Loading branch information
Showing
4 changed files
with
403 additions
and
131 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,137 @@ | ||
| write_attrs_files <- function(dataset, drop_cols, filepath){ | ||
| #' @description generates a text file of column names | ||
| #' | ||
| #' @param dataset the dataset whose columns will be written to a txt file | ||
| #' @param drop_cols any columns to not include in the file. | ||
| #' Passed to select(-all_of(drop_cols)) | ||
| #' @param filepath relative filepath to the txt file | ||
| #' | ||
| #' @return txt file containing the column names of dataset without the drop_columns | ||
|
|
||
| col_names <- select(dataset, -all_of(drop_cols)) %>% | ||
| colnames() %>% | ||
| sort() | ||
|
|
||
| write_csv(x = as.data.frame(col_names), file = filepath, col_names = FALSE) | ||
|
|
||
| return(filepath) | ||
| } | ||
|
|
||
| write_reaches <- function(reach_ids, filepath){ | ||
| #' @description generates a text file of the reach_ids | ||
| #' | ||
| #' @param reach_ids character string of the IDs to be written to a txt file | ||
| #' @param filepath relative filepath to the txt file | ||
| #' | ||
| #' @return txt file containing the reach_ids | ||
|
|
||
| reach_ids <- sort(unique(reach_ids)) | ||
|
|
||
| write_csv(x = as.data.frame(reach_ids), file = filepath, col_names = FALSE) | ||
|
|
||
| return(filepath) | ||
| } | ||
|
|
||
| write_reaches_cv <- function(train_dataset, cv_folds, out_dir){ | ||
| #' @description generates a text file of the reach_ids within the validation set, | ||
| #' one text file per fold. | ||
| #' | ||
| #' @param train_dataset tibble with columns for PRMS_segid and data_type. | ||
| #' This will be split into cv_folds partitions based on the data_type proportions | ||
| #' @param cv_folds number of cross validation folds in training. This is used to | ||
| #' write validation files. | ||
| #' @param out_dir directory to save txt files | ||
| #' | ||
| #' @return txt files containing the reach_ids | ||
|
|
||
| #create validation file names | ||
| val_nums <- seq(1,cv_folds,1) | ||
| filepaths <- file.path(out_dir, c(paste0('spatial_val_reaches', val_nums, '.txt'))) | ||
|
|
||
| #add group variable to dataset based on spatial splits | ||
| train_dataset$group <- 0 | ||
|
|
||
| #determine validation splits and write to file | ||
| for(i in val_nums){ | ||
| if(i == cv_folds){ | ||
| ind_i <- which(train_dataset$group == 0) | ||
| inds_grp <- seq(1,length(ind_i),1) | ||
| }else{ | ||
| #only use data that haven't been assigned to a group | ||
| ind_i <- which(train_dataset$group == 0) | ||
| #get the correct training proportion to use for a reduced dataset | ||
| train_prop_i <- nrow(train_dataset)/cv_folds/nrow(train_dataset[ind_i,]) | ||
| #group row indices | ||
| inds_grp <- make_spatial_split_CVtraining(attrs_df = train_dataset[ind_i,], | ||
| train_prop = train_prop_i) | ||
| } | ||
| #assign group index | ||
| train_dataset$group[ind_i][inds_grp] <- i | ||
| write_csv(x = as.data.frame(sort(unique(train_dataset$PRMS_segid[train_dataset$group == i]))), | ||
| file = filepaths[i], col_names = FALSE) | ||
| } | ||
|
|
||
| return(filepaths) | ||
| } | ||
|
|
||
| write_dates <- function(dataset, cv_folds = NULL, out_dir){ | ||
| #' @description generates a text file for each of the training and testing | ||
| #' start and end dates for use in Python. | ||
| #' | ||
| #' @param dataset list of training and testing datasets, each with a Date column | ||
| #' @param cv_folds number of cross validation folds in training. This is used to | ||
| #' write validation files. | ||
| #' @param out_dir directory to save txt files | ||
| #' | ||
| #' @return vector of paths to txt files containing the start and end dates | ||
|
|
||
| if (is.null(cv_folds)){ | ||
| filenames <- file.path(out_dir, c('train_start.txt', 'train_end.txt', 'test_start.txt', 'test_end.txt')) | ||
| }else{ | ||
| #create validation files from training time period | ||
| val_nums <- seq(1,cv_folds-1,1) | ||
| filenames <- file.path(out_dir, c('train_start.txt', 'train_end.txt', 'test_start.txt', 'test_end.txt', | ||
| paste0('val_start', val_nums+1, '.txt'), paste0('val_end', val_nums, '.txt'))) | ||
| } | ||
|
|
||
| train_start <- min(dataset$training$Date) %>% as.data.frame() | ||
| train_end <- max(dataset$training$Date) %>% as.data.frame() | ||
| test_start <- min(dataset$testing$Date) %>% as.data.frame() | ||
| test_end <- max(dataset$testing$Date) %>% as.data.frame() | ||
|
|
||
| write_csv(train_start, file = filenames[1], col_names = FALSE) | ||
| write_csv(train_end, file = filenames[2], col_names = FALSE) | ||
| write_csv(test_start, file = filenames[3], col_names = FALSE) | ||
| write_csv(test_end, file = filenames[4], col_names = FALSE) | ||
|
|
||
| if (!is.null(cv_folds)){ | ||
| #determine validation start and end dates and write to file | ||
| for(i in val_nums){ | ||
| val_start <- train_start + (train_end - train_start)/cv_folds*i | ||
| val_end <- val_start - 1 | ||
| write_csv(val_start, file = filenames[4+i], col_names = FALSE) | ||
| write_csv(val_end, file = filenames[8+i], col_names = FALSE) | ||
| } | ||
| } | ||
|
|
||
| return(filenames) | ||
| } | ||
|
|
||
| write_pred_obs <- function(pred_obs, out_dir){ | ||
| #' @description generates a text file containing the predictions, observations, | ||
| #' date, and segment columns. | ||
| #' | ||
| #' @param pred_obs dataframe containing columns "Date", "PRMS_segid", "obs", and ".pred" | ||
| #' @param out_dir directory to save txt file | ||
| #' | ||
| #' @return vector of paths to txt file | ||
|
|
||
| filename <- file.path(out_dir, 'pred_obs.txt') | ||
|
|
||
| pred_obs <- select(pred_obs, "Date", "PRMS_segid", "obs", ".pred") %>% | ||
| rename(pred = '.pred') | ||
|
|
||
| write_csv(pred_obs, filename) | ||
|
|
||
| return(filename) | ||
| } |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Oops, something went wrong.