The autumn's discuss from aaronrudkin

autumn's Issues

Flag to collapse small raking targets

It has been brought up that a useful flag for harvest would be the ability to automatically collapse target levels (either to remove smaller levels and improve convergence, or to solve the problem of missing interaction levels). I'm not fully sure how I'd implement this, but I think it's a good idea.

v0.10 checklist

Things to do before v0.10:

Floating point comparison issues with "targets that do not sum to 1" error

Great package - thanks!

I've been playing around with it, and I've encountered an issue (maybe a bug?).

It looks like the function check_any_data_issues is looking to see if all of the weighting targets add up to 1. I have a situation where they do all add up to 1, but I'm getting the error message "Target variable ... has targets that do not sum to 1." I suspect this is a floating point comparison issue.

Here's a very small example that should be reproducible (let me know if it doesn't work for you):

atlanta <- structure(list(w_race = c("White", "White", "White", "White", 
"Black", "White", "Black", "White", "Other race", "Black", "White", 
"White", "White", "White", "White", "White", "White", "White", 
"White", "White")), row.names = c(NA, -20L), class = c("tbl_df", 
"tbl", "data.frame"))

(atlanta)
#>        w_race
#> 1       White
#> 2       White
#> 3       White
#> 4       White
#> 5       Black
#> 6       White
#> 7       Black
#> 8       White
#> 9  Other race
#> 10      Black
#> 11      White
#> 12      White
#> 13      White
#> 14      White
#> 15      White
#> 16      White
#> 17      White
#> 18      White
#> 19      White
#> 20      White

targets <- structure(list(variable = c("w_race", "w_race", "w_race"), level = c("Black", 
"Other race", "White"), proportion = c(0.299944881294484, 0.0993062927185731, 
0.600748825986942)), row.names = c(NA, -3L), class = c("tbl_df", 
"tbl", "data.frame"))

(targets)
#>   variable      level proportion
#> 1   w_race      Black 0.29994488
#> 2   w_race Other race 0.09930629
#> 3   w_race      White 0.60074883

sum(targets$proportion)
#> [1] 1

autumn::harvest(atlanta, target = targets)
#> Error in check_any_data_issues(data, target, weights): Errors detected in weight targets:
#> Target variable `w_race` has targets that do not sum to 1.

^{Created on 2020-07-14 by the reprex package (v0.3.0)}

If I'm missing something, please let me know. Thanks again for all of your work on the package.

Error for missing levels in the data

I have some variables in a data set the a want to weight. However, not all levels are present in the data set.

library(dplyr)
library(autumn)

harvest(d, weights)

So I get this error:

Error in check_any_data_issues(data, target, weights) : Errors detected in data. Some variables have values in the weight targets which are not present in the data:`

Here is a dput of the quotes

dput(weights)
list(Rec_Age = c(`1` = 0, `2` = 0.181, `3` = 0.2877, `4` = 0.3311, 
`5` = 0.2001), Rec_Income = c(`1` = 0.1105, `2` = 0.2852, `3` = 0.2343, 
`4` = 0.3699), Q6 = c(`1` = 0.067, `2` = 0.3409, `3` = 0.592), 
    RECQ5_1 = c(`1` = 0.4099, `2` = 0.5239, `3` = 0.0662), RECQ5_2 = c(`1` = 0.1621, 
    `2` = 0.3803, `3` = 0.4576), RECQ5_3 = c(`1` = 0.0508, `2` = 0.294, 
    `3` = 0.6551), RECQ5_4 = c(`1` = 0.103, `2` = 0.4864, `3` = 0.4106
    ))

and the data:


dput(d)
structure(list(RESPID = structure(c(459, 311, 223, 60, 613, 495, 
300, 273, 78, 170, 217, 61, 175, 619, 270, 218, 453, 492, 23, 
65, 33, 113, 532, 26, 119, 49, 208, 102, 200, 165, 435, 298, 
593, 220, 111, 53, 494, 271, 305, 420, 323, 607, 105, 19, 426, 
171, 330, 201, 332, 277), label = "RESPID - Respondent ID", format.spss = "F10.0", display_width = 0L), 
    Rec_Age = structure(c(4, 2, 4, 3, 4, 4, 4, 3, 2, 2, 3, 2, 
    3, 4, 4, 2, 4, 4, 2, 3, 2, 2, 2, 3, 3, 2, 2, 2, 2, 3, 2, 
    3, 2, 3, 4, 3, 4, 3, 2, 3, 3, 3, 4, 4, 4, 2, 2, 3, 4, 3), label = "Rec_Age - Recode Age", format.spss = "F1.0", display_width = 0L), 
    Rec_Income = structure(c(3, 1, 2, 1, 1, 2, 2, 3, 2, 1, 2, 
    2, 2, 1, 1, 2, 2, 3, 3, 2, 2, 2, 2, 3, 2, 3, 2, 2, 1, 2, 
    2, 2, 1, 3, 1, 1, 1, 1, 1, 3, 3, 2, 3, 3, 2, 2, 2, 2, 2, 
    2), label = "Rec_Income - Recode Income", format.spss = "F1.0", display_width = 0L), 
    Q6 = structure(c(2, 1, 2, 3, 2, 3, 2, 1, 3, 2, 2, 3, 3, 3, 
    2, 2, 3, 3, 2, 1, 2, 3, 3, 2, 2, 2, 1, 2, 1, 2, 2, 3, 3, 
    2, 3, 2, 3, 2, 2, 1, 3, 2, 2, 2, 3, 2, 2, 1, 3, 2), label = "Q6 - Wie stark interessieren Sie sich für Bekleidung und Mode?", format.spss = "F1.0", display_width = 0L), 
    RECQ5_1 = c(1, 1, 2, 2, 2, 1, 1, 1, 2, 1, 2, 2, 2, 2, 1, 
    1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 1, 1, 1, 2, 1, 1, 2, 2, 
    3, 1, 1, 2, 1, 1, 2, 1, 2, 1, 1, 1, 1, 2, 1, 1), RECQ5_2 = c(2, 
    2, 3, 3, 3, 2, 3, 1, 3, 3, 3, 3, 3, 3, 3, 2, 3, 3, 1, 3, 
    2, 3, 2, 2, 2, 1, 1, 1, 1, 2, 2, 3, 3, 3, 3, 3, 3, 2, 1, 
    1, 3, 3, 2, 1, 3, 1, 2, 1, 3, 2), RECQ5_3 = c(3, 1, 3, 3, 
    3, 3, 2, 2, 3, 3, 3, 3, 3, 3, 2, 2, 3, 3, 1, 2, 2, 3, 3, 
    2, 2, 2, 1, 3, 3, 2, 2, 3, 2, 3, 3, 2, 3, 3, 2, 1, 3, 3, 
    3, 2, 3, 1, 3, 3, 3, 2), RECQ5_4 = c(1, 2, 2, 2, 2, 2, 1, 
    1, 3, 2, 2, 3, 3, 3, 1, 1, 2, 3, 1, 1, 1, 3, 2, 1, 2, 1, 
    1, 1, 1, 2, 1, 3, 3, 3, 2, 1, 2, 2, 1, 1, 2, 1, 2, 1, 2, 
    1, 2, 3, 2, 2)), row.names = c(NA, -50L), class = "data.frame")

aaronrudkin / autumn Goto Github PK

autumn's Issues

Flag to collapse small raking targets

v0.10 checklist

Floating point comparison issues with "targets that do not sum to 1" error

Error for missing levels in the data

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent