The example code to check the problem in a question problem, posted on StackOverflow

pandas-mem.py

#!/env/bin/python
# See: http://stackoverflow.com/questions/41893967/increase-in-memory-usage-on-pandas-dataframe-creation/42000635
import sys
import pandas as pd
import gc

# @profile  ## uncomment this line if you want to profile with memory-profiler Python's module
def make_list():
    pd_arr = []
    for i in range(0,10000):
        pd_arr.append([x for x in range(0,1000)])
    return pd_arr

# @profile
def to_profile():
    pd_arr = make_list()
    pd_df = pd.DataFrame.from_records(pd_arr, columns=[x for x in range(0,1000)])
    # pd_df.info(memory_usage='deep')
    print sys.getsizeof(pd_arr), sys.getsizeof(pd_arr[0])
    print sys.getsizeof(pd_df), len(pd_arr)
    print sys.getrefcount(pd_arr)
    del pd_arr
    gc.collect()

to_profile()