search

eliocamp / metR

Tools for Easier Analysis of Meteorological Fields
https://eliocamp.github.io/metR/
140 stars 22 forks source link

Error after na.fill proposes to use na.fill #96

Closed bergmul closed 4 years ago

bergmul commented 5 years ago

Hi,

after using na.fill, I have encountered an error and the package proposes to use na.fill.

Code:

breaks = seq(0,1, by=0.05)
ggplot(norm65,
       aes(x = Cx,
           y = Cy,
           z = A_before_B)) +
  geom_contour_fill(breaks = breaks,
                    na.fill = TRUE)

Resulting Error:

Warning messages:
1: imputing missing values 
2: data must not have missing values. Use na.fill = TRUE or impute them before plotting.

Data:

The data consists of roughly 50 points in a 2 dimensional grid with the z-value being a probability.

Issue:

I stumbled on your package for contour line level labels and tried the contour_fill function to replace my current workflow of first running a LOESS regression on the data to use with geom_raster + geom_contour.

The error is a bit comical and I don't see what's going wrong. Potentially na.fill needs more data to properly extrapolate to the full space resulting in missing values being picked up even after using na.fill? However, if it's user error or a problem with the data, it would be good if the package gives a better error message.

eliocamp commented 5 years ago

No idea what could be 🤔️. Could you give me a minimal reproducible example so I can investigate the error? It probably is data-dependent.

eliocamp commented 5 years ago

Hi, @bergmul , are you still having this issue? Can you provide a reproducible example so I can find and fix the bug? Thanks!

alexmill commented 4 years ago

Hi, here's an MRE based on an example of this bug I am encountering. On

library(metR)

plot_data = data.frame(rbind(
    c(0.803938, 0.0000, 0.00000000),
    c(4.627267, 0.0025, 0.54040403),
    c(4.627267, 0.0100, 0.56553910),
    c(4.627267, 1.0000, 2.80256044),
    c(4.627267, 0.0400, 0.90486256),
    c(4.627267, 0.0900, 1.40756399),
    c(4.627267, 0.1600, 1.45783413),
    c(4.627267, 0.2500, 1.86627904),
    c(4.627267, 0.5625, 2.46952075),
    c(5.530359, 0.0025, 0.27020202),
    c(5.530359, 0.0100, 0.43357998),
    c(5.530359, 1.0000, 3.12931636),
    c(5.530359, 0.0400, 0.96770024),
    c(5.530359, 0.0900, 1.43898283),
    c(5.530359, 0.1600, 1.94168425),
    c(5.530359, 0.2500, 2.52607466),
    c(5.530359, 0.5625, 3.05391115),
    c(6.219067, 0.0025, 0.14452666),
    c(6.219067, 0.0100, 0.47756635),
    c(6.219067, 1.0000, 3.14188390),
    c(6.219067, 0.0400, 1.19391588),
    c(6.219067, 0.0900, 1.94796802),
    c(6.219067, 0.1600, 2.53864219),
    c(6.219067, 0.2500, 2.93451956),
    c(6.219067, 0.5625, 3.12303260),
    c(9.210564, 0.0025, 0.03141884),
    c(9.210564, 0.0100, 1.11536878),
    c(9.210564, 1.0000, 3.14188390),
    c(9.210564, 0.0400, 3.12931636),
    c(9.210564, 0.0900, 3.14188390),
    c(9.210564, 0.1600, 3.14188390),
    c(9.210564, 0.2500, 3.14188390),
    c(9.210564, 0.5625, 3.14188390),
    c(6.909987, 0.0025, 0.12567536),
    c(6.909987, 0.0100, 0.46499882),
    c(6.909987, 1.0000, 3.14188390),
    c(6.909987, 0.0400, 1.58350949),
    c(6.909987, 0.0900, 2.60147987),
    c(6.909987, 0.1600, 3.01620854),
    c(6.909987, 0.2500, 3.12931636),
    c(6.909987, 0.5625, 3.14188390),
    c(7.824939, 0.0025, 0.06912145),
    c(7.824939, 0.0100, 0.61580924),
    c(7.824939, 1.0000, 3.14188390),
    c(7.824939, 0.0400, 2.40668307),
    c(7.824939, 0.0900, 3.07904622),
    c(7.824939, 0.1600, 3.14188390),
    c(7.824939, 0.2500, 3.14188390),
    c(7.824939, 0.5625, 3.14188390),
    c(8.517640, 0.0025, 0.08112113),
    c(8.517640, 0.0100, 0.79489663),
    c(8.517640, 1.0000, 3.14188390),
    c(8.517640, 0.0400, 2.94708710),
    c(8.517640, 0.0900, 3.14188390),
    c(8.517640, 0.1600, 3.14188390),
    c(8.517640, 0.2500, 3.14188390),
    c(8.517640, 0.5625, 3.14188390)
))
colnames(plot_data) = c('x','y','z')

ggplot(plot_data, aes(x,y,z=z)) +
     geom_contour_fill(na.fill=TRUE)
eliocamp commented 4 years ago

Thank you very much. I see that you are using a particularly strange grid. This now works on the development version on github (remotes::install_github("eliocamp/metR")) Let me know if it works for you. 

library(metR)
library(ggplot2)

plot_data = data.frame(rbind(
    c(0.803938, 0.0000, 0.00000000),
    c(4.627267, 0.0025, 0.54040403),
    c(4.627267, 0.0100, 0.56553910),
    c(4.627267, 1.0000, 2.80256044),
    c(4.627267, 0.0400, 0.90486256),
    c(4.627267, 0.0900, 1.40756399),
    c(4.627267, 0.1600, 1.45783413),
    c(4.627267, 0.2500, 1.86627904),
    c(4.627267, 0.5625, 2.46952075),
    c(5.530359, 0.0025, 0.27020202),
    c(5.530359, 0.0100, 0.43357998),
    c(5.530359, 1.0000, 3.12931636),
    c(5.530359, 0.0400, 0.96770024),
    c(5.530359, 0.0900, 1.43898283),
    c(5.530359, 0.1600, 1.94168425),
    c(5.530359, 0.2500, 2.52607466),
    c(5.530359, 0.5625, 3.05391115),
    c(6.219067, 0.0025, 0.14452666),
    c(6.219067, 0.0100, 0.47756635),
    c(6.219067, 1.0000, 3.14188390),
    c(6.219067, 0.0400, 1.19391588),
    c(6.219067, 0.0900, 1.94796802),
    c(6.219067, 0.1600, 2.53864219),
    c(6.219067, 0.2500, 2.93451956),
    c(6.219067, 0.5625, 3.12303260),
    c(9.210564, 0.0025, 0.03141884),
    c(9.210564, 0.0100, 1.11536878),
    c(9.210564, 1.0000, 3.14188390),
    c(9.210564, 0.0400, 3.12931636),
    c(9.210564, 0.0900, 3.14188390),
    c(9.210564, 0.1600, 3.14188390),
    c(9.210564, 0.2500, 3.14188390),
    c(9.210564, 0.5625, 3.14188390),
    c(6.909987, 0.0025, 0.12567536),
    c(6.909987, 0.0100, 0.46499882),
    c(6.909987, 1.0000, 3.14188390),
    c(6.909987, 0.0400, 1.58350949),
    c(6.909987, 0.0900, 2.60147987),
    c(6.909987, 0.1600, 3.01620854),
    c(6.909987, 0.2500, 3.12931636),
    c(6.909987, 0.5625, 3.14188390),
    c(7.824939, 0.0025, 0.06912145),
    c(7.824939, 0.0100, 0.61580924),
    c(7.824939, 1.0000, 3.14188390),
    c(7.824939, 0.0400, 2.40668307),
    c(7.824939, 0.0900, 3.07904622),
    c(7.824939, 0.1600, 3.14188390),
    c(7.824939, 0.2500, 3.14188390),
    c(7.824939, 0.5625, 3.14188390),
    c(8.517640, 0.0025, 0.08112113),
    c(8.517640, 0.0100, 0.79489663),
    c(8.517640, 1.0000, 3.14188390),
    c(8.517640, 0.0400, 2.94708710),
    c(8.517640, 0.0900, 3.14188390),
    c(8.517640, 0.1600, 3.14188390),
    c(8.517640, 0.2500, 3.14188390),
    c(8.517640, 0.5625, 3.14188390)
))
colnames(plot_data) = c('x','y','z')

ggplot(plot_data, aes(x,y,z=z)) +
    geom_contour_fill(na.fill = TRUE) +
    geom_point() 
#> Warning: imputing missing values

Created on 2019-10-04 by the reprex package (v0.3.0)