Help us understand the problem. What is going on with this article?

Kerasの損失関数履歴をデータフレームに変換する

More than 1 year has passed since last update.

Kerasの損失関数の履歴が欲しい

 自分の作ったモデルがどのように収束していくかは非常に大事な情報です。
そこで、Kerasの損失関数(などの)履歴をファイルに残すために、Pandasのデータフレームに変換するスクリプトを紹介します

history_callback = model.fit(x_train, y_train, ....)
loss_history = history_callback.history

これで履歴が記録されます

loss_history #dict型
list_losss_history = list(loss_history.items()) #listへ変換
print(type(loss_history),type(list_losss_history))
print(list_losss_history)

出力

<class 'dict'> <class 'list'>
[('loss', [0.5525253211575696, 0.4036650247471646, 0.36141999579291056, 0.33595747268301807, 0.3177382633604555, 0.29891023940306444, 0.2881778943508099, 0.27792962343876176, 0.26560747584624167, 0.25740950032279025]), ('val_acc', [0.8358461538461538, 0.8772307692307693, 0.8713846153846154, 0.8776923076923077, 0.8813846153846154, 0.8852307692307693, 0.8812307692307693, 0.8816923076923077, 0.8872307692307693, 0.9003076923076923]), ('val_loss', [0.4148623236967967, 0.35430374641487233, 0.3401262025305858, 0.33813549585525804, 0.3276629987571102, 0.30808376993124303, 0.32132247155217025, 0.340628393916843, 0.30153624791250777, 0.2850437596107905]), ('acc', [0.8024273504273505, 0.8550427350427351, 0.8704615384615385, 0.8781880341880342, 0.8857264957264958, 0.8917264957264958, 0.8942222222222223, 0.8978803418803418, 0.9025982905982906, 0.9046666666666666])]

変換するスクリプト

list2D=[]
colname=[]
for i in range(len(list_losss_history)):
    list2D.append(list_losss_history[i][1])#単純にリストに追加していく
    colname.append(list_losss_history[i][0])
hist_data = pd.DataFrame(np.array(list2D).T,columns=colname)
print(hist_data)

出力

       loss   val_acc  val_loss       acc
0  0.552525  0.835846  0.414862  0.802427
1  0.403665  0.877231  0.354304  0.855043
2  0.361420  0.871385  0.340126  0.870462
3  0.335957  0.877692  0.338135  0.878188
4  0.317738  0.881385  0.327663  0.885726
5  0.298910  0.885231  0.308084  0.891726
6  0.288178  0.881231  0.321322  0.894222
7  0.277930  0.881692  0.340628  0.897880
8  0.265607  0.887231  0.301536  0.902598
9  0.257410  0.900308  0.285044  0.904667

あとは、ファイルにセーブすればOKです。

参考URL

https://codeday.me/jp/qa/20190301/346082.html
python – Kerasの損失出力をファイルに記録する方法
Lossのみの結果を取り出す方法が書いてあります

officefutaro
初めて使ったコンピュータはTK-80BSです。 現在はデータ分析を主要な道具とした問題解決で生活しています
https://futaro.hatenadiary.jp/
Why not register and get more from Qiita?
  1. We will deliver articles that match you
    By following users and tags, you can catch up information on technical fields that you are interested in as a whole
  2. you can read useful information later efficiently
    By "stocking" the articles you like, you can search right away
Comments
No comments
Sign up for free and join this conversation.
If you already have a Qiita account
Why do not you register as a user and use Qiita more conveniently?
You need to log in to use this function. Qiita can be used more conveniently after logging in.
You seem to be reading articles frequently this month. Qiita can be used more conveniently after logging in.
  1. We will deliver articles that match you
    By following users and tags, you can catch up information on technical fields that you are interested in as a whole
  2. you can read useful information later efficiently
    By "stocking" the articles you like, you can search right away
ユーザーは見つかりませんでした