Installation Guide

This guide covers all installation options for OTEX, including optional dependencies and data access configuration.

Table of Contents

Requirements

System Requirements

  • Python: 3.9 or higher

  • Operating System: Linux, macOS, or Windows

  • Memory: 4 GB RAM minimum (8 GB recommended for large analyses)

  • Disk Space: 500 MB for installation, additional space for data downloads

Python Dependencies

Core dependencies (installed automatically):

Package

Version

Purpose

numpy

≥1.20

Numerical computing

pandas

≥1.3

Data manipulation

scipy

≥1.7

Scientific computing

matplotlib

≥3.4

Visualization

xarray

≥0.19

N-dimensional arrays

netCDF4

≥1.5

NetCDF file support

tables

≥3.6

HDF5 file support

tqdm

≥4.60

Progress bars

Basic Installation

From GitHub

pip install git+https://github.com/msotocalvo/OTEX.git

Optional Dependencies

SALib (Uncertainty Analysis)

SALib is required for Sobol sensitivity analysis:

pip install otex[uncertainty]

Or install separately:

pip install SALib>=1.4.0

Siting Layers (Geospatial Filtering)

Site-screening for protected areas, shipping lanes, and natural hazards requires geospatial libraries:

pip install otex[siting]

This installs:

Package

Purpose

geopandas ≥0.12

Vector geospatial operations (point-in-polygon for WDPA)

rasterio ≥1.3

Raster sampling (vessel density, PGA)

shapely ≥2.0

Geometry buffering

pyproj ≥3.4

Coordinate reference system reprojection

requests ≥2.28

Layer downloads

The first time a siting layer is needed, OTEX downloads it on demand to ~/.otex/siting_cache/. Total cache size is ~5 GB for the full set (WDPA + vessel density + PGA + IBTrACS). See the Siting tutorial for details.

All Optional Dependencies

pip install otex[all]

This includes:

  • CoolProp

  • SALib

  • geopandas, rasterio, shapely, pyproj, requests (siting)

  • pytest and pytest-cov (for testing)

Development Installation

For contributing to OTEX or modifying the source code:

# Clone the repository
git clone https://github.com/msotocalvo/OTEX.git
cd OTEX

# Create virtual environment (recommended)
python -m venv venv
source venv/bin/activate  # Linux/macOS
# or: venv\Scripts\activate  # Windows

# Install in development mode with all dependencies
pip install -e ".[dev,all]"

# Run tests to verify installation
pytest tests/ -v

Oceanographic Data Access

OTEX supports two oceanographic data sources: CMEMS (Copernicus Marine) and HYCOM (Hybrid Coordinate Ocean Model).

HYCOM Data Access (No Authentication)

HYCOM data is freely available via OPeNDAP with no account required. This is the easiest way to get started:

from otex.regional import run_regional_analysis

otec_plants, sites = run_regional_analysis(
    studied_region='Jamaica',
    data_source='HYCOM',
    year_start=2020,
    year_end=2020,
)

Available HYCOM datasets:

Dataset

Period

Description

GLBv0.08/expt_53.X

1994–2015

Reanalysis

GLBy0.08/expt_93.0

2019–2024

Analysis

Note: HYCOM data is not available for 2016–2018 (gap between experiments). Use CMEMS for those years.

CMEMS Data Access

CMEMS provides a longer continuous time series (1993–present) but requires a free Copernicus Marine account.

Step 1: Create Account

  1. Go to Copernicus Marine

  2. Click “Register” and create an account

  3. Verify your email address

Step 2: Configure Credentials

Option A: Using copernicusmarine CLI (Recommended)

# Install the CLI tool
pip install copernicusmarine

# Login (stores credentials securely)
copernicusmarine login

Follow the prompts to enter your username and password.

Option B: Environment Variables

export COPERNICUSMARINE_SERVICE_USERNAME="your_username"
export COPERNICUSMARINE_SERVICE_PASSWORD="your_password"

Add these to your ~/.bashrc or ~/.zshrc for persistence.

Option C: Configuration File

Create ~/.copernicusmarine/credentials with:

username: your_username
password: your_password

Step 3: Verify Access

from otex.data.cmems import download_data
# If credentials are configured correctly, this won't raise an error

Data Storage

Downloaded data is cached locally in the Data_Results/ directory. Both CMEMS and HYCOM downloads use the same directory structure and CMEMS-compatible file format:

Data_Results/
├── Jamaica/
│   ├── Jamaica_2020_50.0_MW_low_cost/
│   │   ├── T_22.0m_2020_Jamaica.h5      # Warm water temperatures
│   │   ├── T_1062.0m_2020_Jamaica.h5    # Cold water temperatures
│   │   └── OTEC_sites_Jamaica_*.csv     # Results
│   └── T_*m_2020_Jamaica_*.nc           # Raw NetCDF downloads
├── Philippines/
└── ...

Verifying Installation

Basic Verification

# Test import
import otex
print(f"OTEX version: {otex.__version__}")

# Test configuration
from otex.config import parameters_and_constants
inputs = parameters_and_constants()
print(f"Default cycle: {inputs['cycle_type']}")
print(f"Default fluid: {inputs['fluid_type']}")

CoolProp Verification

from otex.core.fluids import get_working_fluid

# This will use CoolProp if available
fluid = get_working_fluid('ammonia', use_coolprop=True)
print(f"Fluid type: {type(fluid).__name__}")

If CoolProp is not installed, you’ll see a warning and polynomial correlations will be used.

Uncertainty Module Verification

from otex.analysis import MonteCarloAnalysis, UncertaintyConfig

config = UncertaintyConfig(n_samples=10, parallel=False)
mc = MonteCarloAnalysis(T_WW=28.0, T_CW=5.0, config=config)
results = mc.run(show_progress=False)
print(f"LCOE samples: {len(results.lcoe)}")

Full Test Suite

# Run all tests
pytest tests/ -v

# Run with coverage
pytest tests/ --cov=otex --cov-report=html

# Skip slow tests
pytest tests/ -v -m "not slow"

Troubleshooting

Common Issues

CoolProp Installation Fails

On some systems, CoolProp requires compilation. Try:

# Ubuntu/Debian
sudo apt-get install python3-dev build-essential

# macOS
xcode-select --install

# Then retry
pip install CoolProp

HDF5/netCDF4 Issues

# Ubuntu/Debian
sudo apt-get install libhdf5-dev libnetcdf-dev

# macOS
brew install hdf5 netcdf

# Then reinstall
pip install --force-reinstall h5py netCDF4

CMEMS Download Errors

  1. Verify credentials:

    copernicusmarine login --check

  2. Check internet connection and firewall settings

  3. Verify data product availability:

    copernicusmarine describe --contains GLOBAL_MULTIYEAR_PHY

  4. Alternative: Try HYCOM instead (no credentials needed):

    run_regional_analysis(studied_region='Jamaica', data_source='HYCOM',
                      year_start=2020, year_end=2020)

HYCOM Download Errors

  1. HYCOM OPeNDAP servers may be temporarily unavailable — retry after a few minutes

  2. Verify the year falls within available ranges (1994–2015 or 2019–2024)

  3. Check internet connection (HYCOM uses port 443 via HTTPS)

Memory Errors

For large analyses, increase available memory or reduce sample size:

config = UncertaintyConfig(n_samples=500)  # Reduce from 1000

Getting Help

Next Steps