TypeError：遍历每一行以获得特定列值时，字符串索引必须是整数

Question

我想对每个基因进行时间线性回归分析，采用 model 中存在的所有变量，因此使用所有基因。

在df5中，x轴代表“基因符号”，y轴代表“时间”。

import pandas as pd
import numpy as np
import plotly.express as px
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import train_test_split

# Split data into training and test splits
train_idx, test_idx = train_test_split(df5.index, test_size=.25, random_state=0)
df5["split"] = "train"
df5.loc[test_idx, "split"] = "test"

# Inputs and targets
X = df5.iloc[:, 1:-1]
y = df5.iloc[:, 0]

X_train = df5.loc[train_idx, ["4", "8", "12", "24", "48"]]
y_train = df5.loc[train_idx, "0"]

# Linear regression prediction
model = LinearRegression()
model.fit(X_train, y_train)
df5['prediction'] = model.predict(X)

当我想使用y=i["prediction"]将y变量设置为每一行的prediction列值时，我得到一个类型错误。

# Scatter plot
for i, j in df5.iterrows():
  for col in df5.columns:
    fig = px.scatter(df5[col], x=df5.iloc[:,0], y=i["prediction"], marginal_x='histogram', marginal_y='histogram', color='split', trendline='ols')
    fig.update_traces(histnorm='probability', selector={'type':'histogram'})
    fig.add_shape(type="line", line=dict(dash='dash'), x0=y.min(), y0=y.min(), x1=y.max(), y1=y.max())
    fig.show()

追溯：

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-395-6ef08290c83a> in <module>()
      2 for i, j in df5.iterrows():
      3   for col in df5.columns:
----> 4     fig = px.scatter(df5[col], x=df5.iloc[:,0], y=i["prediction"], marginal_x='histogram', marginal_y='histogram', color='split', trendline='ols')
      5     fig.update_traces(histnorm='probability', selector={'type':'histogram'})
      6     fig.add_shape(type="line", line=dict(dash='dash'), x0=y.min(), y0=y.min(), x1=y.max(), y1=y.max())

TypeError: string indices must be integers

数据：

df5.head().to_dict()

{'0': {'DNAJB6 /// TMEM135': 0.30131649339447103,
  'DNAJC14': 0.2255444383216058,
  'DNAJC15': 0.25789169794229455,
  'DNAJC30': 0.11388797858763917,
  'DNAJC9': 0.11205541676885071},
 '12': {'DNAJB6 /// TMEM135': 0.28354614480145346,
  'DNAJC14': 0.2343653660720247,
  'DNAJC15': 0.2406210529534205,
  'DNAJC30': 0.11229754447748205,
  'DNAJC9': 0.12045170255898871},
 '24': {'DNAJB6 /// TMEM135': 0.27395808285292367,
  'DNAJC14': 0.2246018336027369,
  'DNAJC15': 0.22347959865906092,
  'DNAJC30': 0.11379897713291527,
  'DNAJC9': 0.10622530623273815},
 '4': {'DNAJB6 /// TMEM135': 0.2949284643966144,
  'DNAJC14': 0.22905481299223704,
  'DNAJC15': 0.22312009403152122,
  'DNAJC30': 0.13114878202076288,
  'DNAJC9': 0.12991396178392187},
 '48': {'DNAJB6 /// TMEM135': 0.289873135093664,
  'DNAJC14': 0.2349502215468218,
  'DNAJC15': 0.17706771640592167,
  'DNAJC30': 0.10857074282633467,
  'DNAJC9': 0.13001391250069522},
 '8': {'DNAJB6 /// TMEM135': 0.2794865791356734,
  'DNAJC14': 0.22228815371920396,
  'DNAJC15': 0.22912018863353348,
  'DNAJC30': 0.11799998627920205,
  'DNAJC9': 0.10520854728987451}}

Answer 1

第一：如果错误告诉你哪一行出了问题，那么首先你可以使用 print()、print(type(...)) 等来检查你在这一行的变量中有什么。

看来您使用了错误的变量。 我认为错误是i["prediction"]因为i应该是index of row ，而不是row with data 。 也许如果您for index, row in df5.iterrow()使用更具可读性的变量，而不是for i,j in df.iterrow()那么您会看到运行index["prediction"]而不是row["prediction"]

---

但坦率地说，我不明白你对 plot 的尝试。

x=df5.iloc[:,0]应该在列中给出所有数据，而不是在行中，但是y=row["prediction"]应该给出一行中的单个值。 这没有道理。 您应该使用y=df5["prediction"]并在没有df5.iterrows()的情况下运行它 - 甚至只使用列名而不是数据px.scatter(df5, x=col, y="prediction", ...)

for col in ["4", "8", "12", "24", "48"]:  # without "0"
    fig = px.scatter(df5, x=col, y="prediction", marginal_x='histogram', marginal_y='histogram', color='split')#, trendline='ols')
    fig.update_traces(histnorm='probability', selector={'type':'histogram'})
    fig.add_shape(type="line", line=dict(dash='dash'), x0=y.min(), y0=y.min(), x1=y.max(), y1=y.max())
    fig.show()

---

完整的工作代码和代码中的示例数据 - 所以每个人都可以简单地复制和运行它

顺便说一句：它在单独的页面上打开每个 plot。 我不得不在scatter中跳过trendline='ols' ，因为它给了我错误ImportError: cannot import name '_centered' from 'scipy.signal.signaltools' (/usr/local/lib/python3.8/dist-packages/scipy/signal/signaltools.py)

data = {'0': {'DNAJB6 /// TMEM135': 0.30131649339447103,
  'DNAJC14': 0.2255444383216058,
  'DNAJC15': 0.25789169794229455,
  'DNAJC30': 0.11388797858763917,
  'DNAJC9': 0.11205541676885071},
 '12': {'DNAJB6 /// TMEM135': 0.28354614480145346,
  'DNAJC14': 0.2343653660720247,
  'DNAJC15': 0.2406210529534205,
  'DNAJC30': 0.11229754447748205,
  'DNAJC9': 0.12045170255898871},
 '24': {'DNAJB6 /// TMEM135': 0.27395808285292367,
  'DNAJC14': 0.2246018336027369,
  'DNAJC15': 0.22347959865906092,
  'DNAJC30': 0.11379897713291527,
  'DNAJC9': 0.10622530623273815},
 '4': {'DNAJB6 /// TMEM135': 0.2949284643966144,
  'DNAJC14': 0.22905481299223704,
  'DNAJC15': 0.22312009403152122,
  'DNAJC30': 0.13114878202076288,
  'DNAJC9': 0.12991396178392187},
 '48': {'DNAJB6 /// TMEM135': 0.289873135093664,
  'DNAJC14': 0.2349502215468218,
  'DNAJC15': 0.17706771640592167,
  'DNAJC30': 0.10857074282633467,
  'DNAJC9': 0.13001391250069522},
 '8': {'DNAJB6 /// TMEM135': 0.2794865791356734,
  'DNAJC14': 0.22228815371920396,
  'DNAJC15': 0.22912018863353348,
  'DNAJC30': 0.11799998627920205,
  'DNAJC9': 0.10520854728987451}
}

import pandas as pd
import numpy as np
import plotly.express as px
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import train_test_split

df5 = pd.DataFrame(data)

# Split data into training and test splits
train_idx, test_idx = train_test_split(df5.index, test_size=.25, random_state=0)
df5["split"] = "train"
df5.loc[test_idx, "split"] = "test"

# Inputs and targets
X = df5.iloc[:, 1:-1]
y = df5.iloc[:, 0]

X_train = df5.loc[train_idx, ["4", "8", "12", "24", "48"]]
y_train = df5.loc[train_idx, "0"]

# Linear regression prediction
model = LinearRegression()
model.fit(X_train, y_train)
df5['prediction'] = model.predict(X)

for col in ["4", "8", "12", "24", "48"]:  # without "0"
    fig = px.scatter(df5, x=col, y="prediction", marginal_x='histogram', marginal_y='histogram', color='split')#, trendline='ols')
    fig.update_traces(histnorm='probability', selector={'type':'histogram'})
    fig.add_shape(type="line", line=dict(dash='dash'), x0=y.min(), y0=y.min(), x1=y.max(), y1=y.max())
    fig.show()

Plot 用于"4"列