辞書の出力のズレを直したい

Question

1万件のツイートを格納したテキストファイルから抽出した名詞を対象に、SlothLibを用いたストップワードの除去と辞書の構築を行いたいと考えています。
下記のように、Jupyterと新規に作成した辞書とで出力が違うのですが、これはどのようなことが起こっていて、どのように修正すればいいのか教えていただけますと幸いです。
宜しくお願いします。

ファイルの中身

Dictionary(0 unique tokens: [])

Jupyter notebook上

Dictionary(2845 unique tokens: ['', '000 ギフト コード', '038', '0458', '100 ガチ恋 距離空間 さっき お誕生日おめでとう']...)

該当のソースコード

# import emoji
#
# words = list(filter(lambda x: x in emoji.UNICODE_EMOJI, words))
#https://qiita.com/chamao/items/7edaba62b120a660657e
from natto import MeCab
import os
import urllib.request
import codecs
from gensim import corpora
from collections import Counter
from collections import defaultdict
frequency = defaultdict(int)

f = open('test40.txt')
corpus = f.read().split("\n")

mecab = MeCab('-d /usr/local/lib/mecab/dic/mecab-ipadic-neologd')

#if tagger.lang == 'ja':

with codecs.open("test40.txt", "r", "utf-8") as f:
    corpus = f.read().split("\n")

rm_list = ["RT","https","co","さん","フォロー","本日","応募","今日","プレゼント","お金","FGO","無料","本人","投稿","動画","ツイート","リツイート","Twitter","ローソン","Peing","http","Amazonギフト券","bot","発売中","Youtube","www","WWW","質問箱","コラボ","フォロワー","DM","いいね","ＲＴ","lawson","://","！","peing","youtube","抽選","jp","リプ","キャンペーン","チケット","期間限定","DHC","日本","amp","人間","チャンネル","配信中","YouTube","WEB","楽しみ","イラスト","くじ","@","__"]

stop_words = []
path = 'stop_words.txt'
with open(path) as g:
    stop_words = g.readlines()

docs = []
for txt in corpus:
    words = mecab.parse(txt, as_nodes=True)
    doc = []

    for w in words:
        if w.feature.split(",")[0] == "名詞":
            if not any(sw in w.surface for sw in stop_words):
                if not any(rm in w.surface for rm in rm_list):
                    if len(w.surface) >= 3:
                        doc.append(str(w.surface))

    doc = ' '.join(doc)
    docs.append(doc)
corpus = docs

str_corpus = str(corpus).split(' ')

dictionary = corpora.Dictionary([corpus])
print(dictionary)

dictionary.filter_extremes(no_below=2, no_above=0.5)
dictionary.save_as_text('test40-dic2.txt')

with open("test40-dic2-3.txt", "a", encoding="utf-8") as h:
    h.write(str(dictionary))

補足情報（FW/ツールのバージョンなど）

iOS 10.12.6, Python 3.7.3, Atom

Answer 1

dictionary.filter_extremes(no_below=2, no_above=0.5)

が

print(dictionary)

の前についていないことに気づきました。