Skip to content

Installation Guide

📦 Required Libraries

This project uses several powerful Python libraries for data analysis, machine learning, and deep learning.

Libraries Overview

  • Pandas

    For handling and manipulating tabular data using DataFrames

  • NumPy

    Provides fast numerical operations and multi-dimensional arrays

  • Matplotlib

    Used to create static, animated, and interactive plots

  • Scikit-learn

    A machine learning library with tools for modeling and evaluation

  • Statsmodels

    Enables statistical analysis and time series exploration

  • TensorFlow

    A powerful library for building and training deep learning models

  • Keras

    High-level API within TensorFlow for fast neural network development

  • Pydot

    Python interface for creating Graphviz DOT graphs (optional)

  • Graphviz

    Graph visualization tools used to render DOT graphs (optional)

🔧 Installation Methods

If you have Anaconda or Miniconda installed, this is the most straightforward method:

conda install -c conda-forge pandas numpy matplotlib scikit-learn statsmodels tensorflow pydot graphviz -y

Why Conda?

Conda handles complex dependencies better, especially for TensorFlow and its GPU support.

Method 2: Using pip

For those using standard Python installations:

# Upgrade pip first
pip install --upgrade pip

# Install core packages
pip install pandas numpy matplotlib scikit-learn statsmodels

# Install TensorFlow (CPU version)
pip install tensorflow

# Optional: For GPU support
pip install tensorflow-gpu

# Optional: For neural network visualization
pip install pydot graphviz

Method 3: Using Requirements File

Create a requirements.txt file:

pandas>=1.3.0
numpy>=1.21.0
matplotlib>=3.4.0
scikit-learn>=1.0.0
statsmodels>=0.13.0
tensorflow>=2.10.0
pydot>=1.4.0
graphviz>=0.20.0

Then install all at once:

pip install -r requirements.txt

🐍 Setting Up Virtual Environment

Best Practice

Always use a virtual environment to avoid package conflicts!

Windows

# Create virtual environment
python -m venv hydro_env

# Activate it
hydro_env\Scripts\activate

# Install packages
pip install -r requirements.txt

# To deactivate when done
deactivate

macOS/Linux

# Create virtual environment
python3 -m venv hydro_env

# Activate it
source hydro_env/bin/activate

# Install packages
pip install -r requirements.txt

# To deactivate when done
deactivate

✅ Verify Installation

After installation, verify everything is working:

import sys
print(f"Python version: {sys.version}")

# Test imports
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import sklearn
import statsmodels
import tensorflow as tf
from tensorflow import keras

# Print versions
print(f"Pandas: {pd.__version__}")
print(f"NumPy: {np.__version__}")
print(f"Scikit-learn: {sklearn.__version__}")
print(f"TensorFlow: {tf.__version__}")
print(f"Keras: {keras.__version__}")

# Test TensorFlow
print(f"TensorFlow GPU Available: {tf.config.list_physical_devices('GPU')}")

Expected output: 

Python version: 3.9.x (or higher)
Pandas: 1.3.x
NumPy: 1.21.x
Scikit-learn: 1.0.x
TensorFlow: 2.10.x
Keras: 2.10.x
TensorFlow GPU Available: [] (or list of GPUs if available)

🚨 Troubleshooting

Common Issues and Solutions

ImportError: No module named 'tensorflow'

Solution: Ensure you've activated your virtual environment and installed TensorFlow: 

pip install tensorflow

Graphviz not found

Solution: Graphviz requires system installation:

Windows: Download from Graphviz website

macOS: 

brew install graphviz

Linux: 

sudo apt-get install graphviz  # Ubuntu/Debian
sudo yum install graphviz       # RHEL/CentOS

Memory errors with large datasets

Solution: Consider using: - Smaller batch sizes in neural networks - Data chunking with pandas - Google Colab for free GPU access

💻 Alternative: Google Colab

If you prefer not to install locally, use Google Colab:

  1. Go to Google Colab
  2. Create a new notebook
  3. Most libraries are pre-installed
  4. For additional packages: 
    !pip install statsmodels pydot

🎯 Next Steps

Now that you have all the required libraries installed, proceed to:

GPU Support

For faster neural network training, consider setting up GPU support: - NVIDIA GPU with CUDA support - Install CUDA toolkit and cuDNN - Install tensorflow-gpu instead of tensorflow

Home

Data Import