英文:
Convert rows into columns based on values in another column
问题
我有一个类似这样的pandas数据框架：
train val score
1 0.6125 0.0827 loss
2 0.8565 0.9845 precision
3 0.7596 0.982 recall
4 0.0466 0.0454 loss
5 0.9897 0.9949 precision
6 0.9884 0.9949 recall
我想将它转换成类似这样的格式：
train_loss train_precision train_recall val_loss val_precision val_recall
1 0.6125 0.8565 0.7596 0.0827 0.9845 0.982
2 0.0466 0.9897 0.9884 0.0454 0.9949 0.9949
英文:
I have a pandas data frame that looks like this
train val score
1 0.6125 0.0827 loss
2 0.8565 0.9845 precision
3 0.7596 0.982 recall
4 0.0466 0.0454 loss
5 0.9897 0.9949 precision
6 0.9884 0.9949 recall
and I want to convert it to something like this
train_loss train_precision train_recall val_loss val_precision val_recall
1 0.6125 0.8565 0.7596 0.0827 0.9845 0.982
2 0.0466 0.9897 0.9884 0.0454 0.9949 0.9949
答案1
得分: 2
你可以使用类似以下的代码：
# 转置一个 pandas 数据框
import pandas as pd
# 创建一个数据框
df = pd.DataFrame({'train': [0.6125, 0.8565, 0.7596, 0.0466, 0.9897, 0.9884], 'val': [0.0827, 0.9845, 0.982, 0.0454, 0.9949, 0.9949], 'score': ['loss', 'precision', 'recall']*2})
# 在 score 列中找到唯一值的列表
new_cols = {f'{i}_{j}': [] for i in ['train', 'val'] for j in df.score.unique()}
# 遍历数据框并将值附加到 new_cols 字典中
for _, row in df.iterrows():
new_cols[f'train_{row["score"]}'].append(row[0])
new_cols[f'val_{row["score"]}'].append(row[1])
# 从 new_cols 字典创建一个新的数据框
new_df = pd.DataFrame(new_cols)
print(new_df)
这段代码返回请求的数据框。
train_loss train_precision train_recall val_loss val_precision val_recall
0 0.6125 0.8565 0.7596 0.0827 0.9845 0.9820
1 0.0466 0.9897 0.9884 0.0454 0.9949 0.9949
英文:
You can use something like the following:
# Transpose a pandas dataframe
import pandas as pd
# Create a dataframe
df = pd.DataFrame({'train': [0.6125, 0.8565, 0.7596, 0.0466, 0.9897, 0.9884], 'val': [0.0827, 0.9845, 0.982, 0.0454, 0.9949, 0.9949], 'score': ['loss', 'precision', 'recall']*2})
# Find list of unique values in the score column
new_cols = {f'{i}_{j}': [] for i in ['train', 'val'] for j in df.score.unique()}
# Iterate over the dataframe and append values to the new_cols dictionary
for _, row in df.iterrows():
new_cols[f'train_{row["score"]}'].append(row[0])
new_cols[f'val_{row["score"]}'].append(row[1])
# Create a new dataframe from the new_cols dictionary
new_df = pd.DataFrame(new_cols)
print(new_df)
This code returns the requested df.
train_loss train_precision train_recall val_loss val_precision val_recall
0 0.6125 0.8565 0.7596 0.0827 0.9845 0.9820
1 0.0466 0.9897 0.9884 0.0454 0.9949 0.9949
答案2
得分: 0
def transform(dataframe):
train_loss, train_precision, train_recall, val_loss, val_precision, val_recall = ([] for i in range(6))
for idx in dataframe.index:
if dataframe['score'][idx] == 'loss':
train_loss.append(dataframe['train'][idx])
val_loss.append(dataframe['val'][idx])
if dataframe['score'][idx] == 'precision':
train_precision.append(dataframe['train'][idx])
val_precision.append(dataframe['val'][idx])
if dataframe['score'][idx] == 'recall':
train_recall.append(dataframe['train'][idx])
val_recall.append(dataframe['val'][idx])
return train_loss, train_precision, train_recall, val_loss, val_precision, val_recall
df = pd.DataFrame()
df['train_loss'], df['train_precision'], df['train_recall'], df['val_loss'], df['val_precision'], df['val_recall'] = transform(dataframe)
英文:
def transform(dataframe):
train_loss, train_precision, train_recall, val_loss, val_precision, val_recall = ([] for i in range(6))
for idx in dataframe.index:
if dataframe['score'][idx] == 'loss':
train_loss.append(dataframe['train'][idx])
val_loss.append(dataframe['val'][idx])
if dataframe['score'][idx] == 'precision':
train_precision.append(dataframe['train'][idx])
val_precision.append(dataframe['val'][idx])
if dataframe['score'][idx] == 'recall':
train_recall.append(dataframe['train'][idx])
val_recall.append(dataframe['val'][idx])
return train_loss, train_precision, train_recall, val_loss, val_precision, val_recall
df = pd.DataFrame()
df['train_loss'], df['train_precision'], df['train_recall'], df['val_loss'], df['val_precision'], df['val_recall'] = transform(dataframe)
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库，让每个人都能够通过互相帮助和分享经验来进步。
评论