Stock Price Prediction using Machine Learning in Python

https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/
Admin
This project demonstrates how to predict stock price movements using machine learning techniques in...
Tokens:3,326
Snippets:45
Trust Score:10
Update:8 months ago
Show doc for...
Tokens:
Raw
### Display First Few Rows of Stock Data (Python)

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Shows the first few rows of the DataFrame to get a quick look at the data structure and content. This is a common first step in data exploration.

```python
df.head()
```

--------------------------------

### Split and Normalize Data for Model Training

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Prepares the data for machine learning model training by selecting features and target variables, normalizing the features using StandardScaler, and then splitting the data into training and validation sets. A test size of 10% and a random state for reproducibility are used.

```python
from sklearn.preprocessing import StandardScaler
from sklearn.model_selection import train_test_split

features = df[['open-close', 'low-high', 'is_quarter_end']]
target = df['target']

scaler = StandardScaler()
features = scaler.fit_transform(features)

X_train, X_valid, Y_train, Y_valid = train_test_split(
    features, target, test_size=0.1, random_state=2022)
print(X_train.shape, X_valid.shape)
```

--------------------------------

### Initialize Correlation Heatmap Figure (Python)

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Sets up a figure for plotting a heatmap to visualize correlations between features. This is a standard step before calculating and displaying feature correlations.

```python
plt.figure(figsize=(10, 10)) 

# As our concern is with the highly 
```

--------------------------------

### Importing Libraries for Machine Learning in Python

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Imports necessary Python libraries for data manipulation, visualization, and machine learning tasks. Includes Pandas for dataframes, NumPy for numerical operations, Matplotlib/Seaborn for plotting, Sklearn for ML algorithms, and XGBoost for gradient boosting.

```python
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sb

from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.linear_model import LogisticRegression
from sklearn.svm import SVC
from xgboost import XGBClassifier
from sklearn import metrics

import warnings
warnings.filterwarnings('ignore')
```

--------------------------------

### Plot Distribution of Stock Features (Python)

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Generates distribution plots for 'Open', 'High', 'Low', 'Close', and 'Volume' features. This helps understand the distribution and identify potential skewness or multi-modality.

```python
features = ['Open', 'High', 'Low', 'Close', 'Volume']

plt.subplots(figsize=(20,10))

for i, col in enumerate(features):
  plt.subplot(2,3,i+1)
  sb.distplot(df[col])
plt.show()
```

--------------------------------

### Plot Pie Chart of Target Variable Distribution (Python)

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Visualizes the distribution of the 'target' variable using a pie chart. This helps assess whether the target variable is balanced, which is important for model training.

```python
plt.pie(df['target'].value_counts().values, 
        labels=[0, 1], autopct='%1.1f%%')
plt.show()
```

--------------------------------

### Describing Dataset Statistics in Python

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Generates descriptive statistics for the numerical columns in the Tesla stock price DataFrame. This includes count, mean, standard deviation, minimum, maximum, and quartile values, offering insights into data distribution.

```python
df.describe()
```

--------------------------------

### Displaying Dataset Information in Python

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Prints a concise summary of the Tesla stock price DataFrame, including the index dtype and columns, non-null values, and memory usage. This is useful for understanding data types and identifying missing values.

```python
df.info()
```

--------------------------------

### Loading Tesla Stock Dataset in Python

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Loads the Tesla stock price dataset from a CSV file into a Pandas DataFrame. It then displays the first five rows to give an initial look at the data, which includes OHLC prices and volume.

```python
df = pd.read_csv('/content/Tesla.csv')
df.head()
```

--------------------------------

### Create New Features and Target Variable (Python)

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Calculates new features like 'open-close' difference and 'low-high' difference. It also creates a 'target' variable indicating if the next day's closing price is higher than the current day's.

```python
df['open-close']  = df['Open'] - df['Close']
df['low-high']  = df['Low'] - df['High']
df['target'] = np.where(df['Close'].shift(-1) > df['Close'], 1, 0)
```

--------------------------------

### Plot Confusion Matrix for Validation Data

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Visualizes the confusion matrix for the first trained model (Logistic Regression) on the validation dataset. This provides a detailed breakdown of prediction accuracy, including true positives, true negatives, false positives, and false negatives.

```python
from sklearn.metrics import ConfusionMatrixDisplay
import matplotlib.pyplot as plt

ConfusionMatrixDisplay.from_estimator(models[0], X_valid, Y_valid)
plt.show()
```

--------------------------------

### Visualize Feature Correlation Heatmap

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Generates a heatmap to visualize correlations between features in the dataset. It focuses on correlations greater than 0.9 to identify highly correlated features. This helps in understanding feature relationships before model development.

```python
import seaborn as sb
import matplotlib.pyplot as plt

sb.heatmap(df.drop('Date', axis=1).corr() > 0.9, annot=True, cbar=False)
plt.show()
```

--------------------------------

### Check for Matching 'Close' and 'Adj Close' Rows (Python)

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Compares the 'Close' and 'Adj Close' columns to determine if they contain identical data across all rows. This helps identify redundant columns.

```python
df[df['Close'] == df['Adj Close']].shape
```

--------------------------------

### Create 'is_quarter_end' Feature (Python)

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Adds a binary feature 'is_quarter_end' which is 1 if the month is a multiple of 3 (indicating a quarter end), and 0 otherwise. This feature can capture seasonal effects related to quarterly reporting.

```python
df['is_quarter_end'] = np.where(df['month']%3==0,1,0)
df.head()
```

--------------------------------

### Plot Tesla Closing Stock Price Over Time (Python)

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Visualizes the closing price of Tesla stocks over a period to identify trends. This plot helps in understanding the stock's historical performance.

```python
plt.figure(figsize=(15,5))
plt.plot(df['Close'])
plt.title('Tesla Close price.', fontsize=15)
plt.ylabel('Price in dollars.')
plt.show()
```

--------------------------------

### Plot Yearly Average Stock Prices (Python)

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Generates bar plots showing the average 'Open', 'High', 'Low', and 'Close' prices for each year. This visualization helps in observing long-term trends in stock prices.

```python
data_grouped = df.drop('Date', axis=1).groupby('year').mean()
plt.subplots(figsize=(20,10))

for i, col in enumerate(['Open', 'High', 'Low', 'Close']):
  plt.subplot(2,2,i+1)
  data_grouped[col].plot.bar()
plt.show()

# This code is modified by Susobhan Akhuli
```

--------------------------------

### Train and Evaluate Machine Learning Models

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Trains three different machine learning models (Logistic Regression, SVC, XGBClassifier) on the prepared data and evaluates their performance using the ROC-AUC score on both training and validation sets. This helps in comparing model effectiveness and identifying potential overfitting.

```python
from sklearn.linear_model import LogisticRegression
from sklearn.svm import SVC
from xgboost import XGBClassifier
from sklearn import metrics

models = [LogisticRegression(), SVC(
  kernel='poly', probability=True), XGBClassifier()]

for i in range(3):
  models[i].fit(X_train, Y_train)

  print(f'{models[i]} : ')
  print('Training Accuracy : ', metrics.roc_auc_score(
    Y_train, models[i].predict_proba(X_train)[:,1]))
  print('Validation Accuracy : ', metrics.roc_auc_score(
    Y_valid, models[i].predict_proba(X_valid)[:,1]))
  print()
```

--------------------------------

### Plot Boxplots of Stock Features (Python)

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Generates boxplots for 'Open', 'High', 'Low', 'Close', and 'Volume' features to visualize data distribution and detect outliers. This helps in identifying data points that deviate significantly from the rest.

```python
plt.subplots(figsize=(20,10))
for i, col in enumerate(features):
  plt.subplot(2,3,i+1)
  sb.boxplot(df[col])
plt.show()
```

--------------------------------

### Group Data by Quarter End and Calculate Mean (Python)

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Groups the DataFrame by the 'is_quarter_end' feature and calculates the mean for other columns. This analysis helps compare stock performance metrics during quarter-end months versus non-quarter-end months.

```python
df.drop('Date', axis=1).groupby('is_quarter_end').mean()
```

--------------------------------

### Displaying Dataset Shape in Python

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Calculates and displays the number of rows and columns in the Tesla stock price DataFrame. This provides a quick overview of the dataset's dimensions, indicating the total number of records and features available.

```python
df.shape
```

--------------------------------

### Check for Null Values in DataFrame (Python)

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Calculates and displays the sum of null values for each column in the DataFrame. This is crucial for data cleaning and preprocessing.

```python
df.isnull().sum()
```

--------------------------------

### Extract Day, Month, Year from Date Column (Python)

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Derives 'day', 'month', and 'year' features from the 'Date' column by splitting the date string. This is a common feature engineering step for time-series data.

```python
splitted = df['Date'].str.split('/', expand=True)

df['day'] = splitted[1].astype('int')
df['month'] = splitted[0].astype('int')
df['year'] = splitted[2].astype('int')

df.head()
```

--------------------------------

### Drop Redundant 'Adj Close' Column (Python)

Source: https://www.geeksforgeeks.org/machine-learning/stock-price-prediction-using-machine-learning-in-python/index

Removes the 'Adj Close' column from the DataFrame as it was found to be redundant with the 'Close' column. This simplifies the dataset for further analysis.

```python
df = df.drop(['Adj Close'], axis=1)
```
Tokens:
Raw