modify the UI to use yaml

rabutler commented 6 years ago

use yaml::yaml.load_files("tst.yml") instead of providing the UI in the R script

example yaml file:

--- # define the CRSS processing for this month's run
start_month: "January"
scenarios:
  - name: "January Official"
    scenarios: ????? [how to specify 35 folders]
    ic:
        powell: 3625
        mead: 1075
    start_year: 2019
    std_ind_plots: TRUE
    shortage_conditions: TRUE
    pe_scatter: TRUE
  - name: August Official
    ic:
        powell: 3627.55
        mead: "C:/alan/mead.csv"
    start_year: 2018
ofolder: "C:/alan"
plot_groups: 
    - "January Official", "August Official"
        end_plot_year: 2060
    - "January Official", "January Stress Test"
        end_plot_year: 2026
   - "August Official", "August Stress Test"
process_data:
    sys_cond_data: TRUE
    pe_data: TRUE
    crss_short_cond_data: FALSE

need to figure out if we specify some parameters for each figure, or for each plot group; need to think about which result in more copies of the same info.

try - each plot group has a list of all the plots to create, and then custom values are specified there.
~alternatively, specify in create_figures - but instead of TRUE/FALSE, each contains multiple entries of scenario groups, and then custom values within that group.~

rabutler-usbr commented 4 years ago

most important

[x] move heatmap to plot_group comparison plots
[x] move pe_clouds to plot_group comparison plots
[x] move 5yr simple to plot_group comparison plots
[x] move short_conditions to create_scenarios individual plots
[x] move pe_scatter_fig to create_scenarios individual plots
[x] document the structure of the list only plot_group left. See https://github.com/BoulderCodeHub/Process-CRSS-Res/wiki/Plotting-Specification-Data-Structures
[x] figure out how rmarkdown processes "r mean(c(1,4,7))" and evaluates it (#100)
[x] convert input from test.yml (via parse_yaml_input.R) into same data structure as the current ui.R files (#100)

important

[x] specify colors
[x] ~specify the "main scenario", and check that only one scenario has the "main" attribute~ Don't need to do this, as we now just say which plots should be created for individual scenarios.
[x] plot_groups #83
[x] pdf_name
[x] annText likely needs to be specified by plot_group
[x] move defaults$annText and defaults$end_year to the plot options
[x] specify defaults - at top level, at plot level, or not at all. If not at all, use pre-specified defaults, if at plot level, those overwrite top level and defaults, and if at top level those overwrite pre-specified.
[x] ensure scenario names are unique
[x] only check that scenario name exists if plots are being created
[x] specified plot_group scenarios, and other specified scenario names should have not duplicates
[x] in the plot_group list, user must specify create = TRUE for the different types. This should not be required in the yaml. Ex: if heat is not NULL, then create should be set to TRUE.

optional parameter:

[x] if prcess_data = TRUE or FALSE, all 4 subs are T/F. Otherwise, let each one be specified manually.
[x] folder$output (defaults to $CRSS_DIR)
[x] folder$extra_label (defaults to '')
[x] extra_label
[x] icDimNumber
[x] legendWrap
[x] colorLabel
[x] specify which scenarios are used as the first and second rows in the simplified 5-year table
[x] std_ind_plots (assumed FALSE) if not specified
[x] shortage_conditions (assumed FALSE) if not specified
[x] pe_scatter (assumed FALSE) if not specified
[x] plot_groups$name
[x] defaults$ic_dim_number
[x] defaults$legend_wrap
[x] defaults$color_label probably should move this to each plot type
[x] get_data parameter. if true, it sets sys_cond_data, pe_data, csd_data, and crss_short_cond_data

rabutler-usbr commented 4 years ago

checks

[x] scenarios are fully specified in the yaml
[x] all scenario names are unique
[x] ~main scenario is specified only once~
[x] mead_pe_scatter$model should either be MTOM or CRSS
[x] checks for scenario names should only occur if the plot is going to be created

BoulderCodeHub / Process-CRSS-Res

modify the UI to use yaml #54