Labels don't change when using metric units for some plots

[DaveEslinger]

Plotting units do not change when metric units are used for Water Quality variables (temp & sonde depth), at least. Still running other plots to test. See image; should be labeled as degree C. Values are correct, labels are not.

[padilla410]

Hi Dave,

Could you please provide a reproducible example? I just tried this with the elksmwq data set and was able to produce a y-axis in fahrenheit and celsius (you must set converted = T for english and converted = F for metric).

[DaveEslinger]

Hi Julie,

Yes, that is what I did for the conversion, which worked on the data, but not the labels. Attached are the raw_boxplot_elknmwq_temp_yr.png file showning the error with elkhorn trtaining data and the Reserve_Level_Plotting_Variables excel file used as input. Only change has been to set converted = F for temperature, depth, and precipitation.

Be sure to check the water quality parameters. They were the ones I found messed up. Checking others now with elk_training

Reserve_Level_Plotting_Variables.xlsx

[DaveEslinger]

Labels in met data seem to be working correctly:

[swmpkim]

I think the problem is in the status report script 01-02_all_plot_analyses.R, line 40:

converted_par <- ifelse(convert_temp && par_inst[j] == 'atemp' || par_inst[j] == 'temp' , T, ifelse(convert_depth && par_inst[j] == 'depth', T, F))

Words are failing me at the moment, but the ifelse evaluates to true even if convert_temp is false, if one of the params is temp:

> par_inst <- c("temp", "depth")
> convert_temp = FALSE
> convert_depth = TRUE
> j <- 1
> par_inst[j]
[1] "temp"
> convert_temp
[1] FALSE
> # want converted_par to be false too
> converted_par <- ifelse(convert_temp && par_inst[j] == 'atemp' || par_inst[j] == 'temp'
+                             , T, ifelse(convert_depth && par_inst[j] == 'depth', T, F))
> converted_par
[1] TRUE
> convert_temp && par_inst[j] == 'atemp'
[1] FALSE
> par_inst[j] == 'atemp' || par_inst[j] == 'temp'
[1] TRUE
> convert_temp && par_inst[j] == 'atemp' || par_inst[j] == 'temp'
[1] TRUE

edited many times because I don't know how to show multiple lines of code, apparently

[swmpkim]

A bit more context for my comment above - converted_par is used in the functions that generate the plot (e.g. raw_boxplot, below), so it needs to be true or false to match what we "want" to see. And it's not evaluating quite right, as in my earlier comment.

raw_box <- raw_boxplot(dat
                           , param = par_inst[j]
                           , target_yr = target_year
                           , criteria = par_threshold[[j]]
                           , log_trans = par_log_transform[j]
                           , converted = converted_par
                           , plot_title = include_station_ttl
                           , season_grps = par_season_grp
                           , season_names = par_season_names
                           , season_start = par_season_start[[1]])

[DaveEslinger]

Thanks for the detective work, Kim!

That looks like there might be a set of missing parenthesis: converted_par <- ifelse(convert_temp && (par_inst[j] == 'atemp' || par_inst[j] == 'temp') , T, ifelse(convert_depth && par_inst[j] == 'depth', T, F))

I'll check that.

[DaveEslinger]

Indeed, that seems to be the issue. Tripped up by order of operations!

If anyone would check, I would appreciate it.

Add the parenthesis in bold below to line 40 in the status report script 01-02_all_plot_analyses.R, line 40:

converted_par <- ifelse(convert_temp && ( par_inst[j] == 'atemp' || par_inst[j] == 'temp' ) , T, ifelse(convert_depth && par_inst[j] == 'depth', T, F))

If that works for others, I'll update the repo code. Thanks for the help!

[swmpkim]

Dave, that works for me in the console! Haven't tried running the full status report yet, but it's giving me the right result interactively.

[DaveEslinger]

This is fixed in the status report script 01-02_all_plot_analyses.R script that is part of a new repository for the scripts that use SWMPrExtension to produce the status reports. Please enter similar issues in that repo: https://github.com/NOAA-OCM/SWMP_Status_Reports