r - melt function applied to json input replaces some values with factors -

i have following json file:

{   "observations": [     {       "type": "ob_type_1",       "data": {         "dynamic": {           "sensor": [             {               "timestamp": 552625694285098,               "a": 1.4921862,               "b": 8.613739             },             {               "timestamp": 552625699285098,               "a": 0.6907272,               "b": 7.6243353             }           ]         },         "static": {           "class1": "abc",           "class2": "xyz"         }       }     }   ] } 

i import using rjson:

library(rjson) raw_json <- fromjson(file=json_file) 

to reproduce please use:

raw_json <-    structure(list(observations = list(structure(list(type = "ob_type_1",          data = structure(list(dynamic = structure(list(sensor = list(             structure(list(timestamp = 552625694285098, = 1.4921862,                  b = 8.613739), .names = c("timestamp", "a", "b")),              structure(list(timestamp = 552625699285098, = 0.6907272,                  b = 7.6243353), .names = c("timestamp", "a", "b")))), .names = "sensor"),              static = structure(list(class1 = "abc", class2 = "xyz"), .names = c("class1",              "class2"))), .names = c("dynamic", "static"))), .names = c("type",      "data")))), .names = "observations") 

i melt using reshape2:

library(reshape2)     reshape_json <- melt(raw_json) 

the value column displays '1' , '2' 'class1' , 'class2'. these 'abc' , 'xyz'. how can achieve this? i've experimented jsonlite in past struggled correct output when json file had nested arrays.

            value        l7 l6     l5      l4   l3 l2           l1 1       ob_type_1      <na> na   <na>    <na> type  1 observations 2 552625694285098 timestamp  1 sensor dynamic data  1 observations 3       1.4921862          1 sensor dynamic data  1 observations 4        8.613739         b  1 sensor dynamic data  1 observations 5 552625699285098 timestamp  2 sensor dynamic data  1 observations 6       0.6907272          2 sensor dynamic data  1 observations 7       7.6243353         b  2 sensor dynamic data  1 observations 8               1      <na> na class1  static data  1 observations 9               2      <na> na class2  static data  1 observations 

options(stringsasfactors = false)