R - "Subset the data" Error: ‘>’ not meaningful for factors -

i have subset data onto people completed survey, variable "disposition"

> names(df) [1] "caseid"      "disposition" "regstate"    "pid7"  "ideo5" "birthyr"                   "gender"      "race"        "educ"        > summary.default(df)         length class  mode    caseid      708    -none- numeric disposition 708    factor numeric regstate    708    factor numeric pid7        708    factor numeric ideo5       708    factor numeric birthyr     708    -none- numeric gender      708    factor numeric race        708    factor numeric educ        708    factor numeric 

now subset data:

disposition <- df$disposition 

i can see complete surveys between 33 , 708

completesurveys <- disposition[33:708] 

i trie select data in following way:

selectdata <- complete.cases(df$caseid, df$regstate, df$pid7, df$ideo, df$birthyr, df$gender, df$race, df$educ) 

and define data surveys completed:

completesurveysdat <- (selectdata & (df$disposition > 32 & df$disposition < 709)) 

unfortunately have got:

warning messages:
1: in ops.factor(df$disposition, 32) : ‘>’ not meaningful factors
2: in ops.factor(df$disposition, 709) : ‘<’ not meaningful factors

i recommend use either data.table or dplyr packages manipulate databases these problems easier.

also, true given filtering factor, won't possible use numeric operators (or not of them). should use as.numeric() function address this.

with data.table:

library(data.table) df <- as.data.table df[as.numeric(disposition)> 33 & as.numeric(disposition)<709] 

with dplyr:

library(dplyr) df <- as.tbl(df) df %>% filter(as.numeric(disposition)> 33,as.numeric(disposition)<709) 

you should type ?dplyr or ?data.table obtain more info, these packages have proven extremely useful manipulate databases.

your output means after filtering got 0 observations, hence as.numeric() not solving problem of factors. recommended using as.numeric(as.character()). make sure first have numeric values , use code showed above.

hope helps