### Install metdig from source

Source: https://github.com/nmcdev/metdig/blob/main/README.md

Clone the metdig repository and install it locally using setup.py. This method is useful for development or when installing from a downloaded archive.

```bash
git clone --recursive https://github.com/nmcdev/metdig.git
cd metdig
python setup.py install
```

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

### Install metdig from GitHub repository

Source: https://github.com/nmcdev/metdig/blob/main/README.md

Install metdig directly from its GitHub repository using pip. This command clones the repository and installs the package.

```bash
pip install git+git://github.com/nmcdev/metdig.git
```

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

### Install Cartopy and metdig via pip

Source: https://github.com/nmcdev/metdig/blob/main/README.md

Use this command to install Cartopy from conda-forge, followed by metdig via pip. Ensure you are in a suitable conda environment.

```bash
conda install -c conda-forge cartopy=0.19.0
pip install metdig
```

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

### Extract Vertical Cross-Section Data

Source: https://context7.com/nmcdev/metdig/llms.txt

Extracts vertical cross-section data along a specified path using 3D grid data. Requires defining start and end points as [lat, lon] coordinates.

```python
import metdig.cal as mdgcal
import metdig.io as mdgio
from datetime import datetime

init_time = datetime(2023, 7, 15, 8)
levels = [1000, 925, 850, 700, 500, 300, 200]

# Read 3D data
rh = mdgio.get_model_3D_grid(
    data_source='cassandra', data_name='ecmwf',
    init_time=init_time, fhour=24,
    var_name='rh', levels=levels, extent=[100, 140, 20, 50]
)
tmp = mdgio.get_model_3D_grid(
    data_source='cassandra', data_name='ecmwf',
    init_time=init_time, fhour=24,
    var_name='tmp', levels=levels, extent=[100, 140, 20, 50]
)

# Define cross-section path
st_point = [30, 110]   # Start point [lat, lon]
ed_point = [40, 130]   # End point [lat, lon]

# Extract cross-section
cross_rh = mdgcal.cross_section(rh, st_point, ed_point)
cross_tmp = mdgcal.cross_section(tmp, st_point, ed_point)
print(f"Cross-section shape: {cross_rh.shape}")
```

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

### Initialize Data Source Connections

Source: https://context7.com/nmcdev/metdig/llms.txt

Configure connections to various meteorological data services including CMADaaS, MICAPS, THREDDS, and Copernicus CDS. Ensure all required credentials and server details are provided.

```python
import metdig.io as mdgio

# Initialize data source configurations
mdgio.config_init(
    CMADaaS_DNS='xxx.xxx.xxx.xxx',        # CMADaaS server IP
    CMADaaS_PORT='xxx',                    # CMADaaS port
    CMADaaS_USER_ID='your_user_id',        # CMADaaS user ID
    CMADaaS_PASSWORD='your_password',      # CMADaaS password
    CMADaaS_serviceNodeId='node_id',       # CMADaaS service node ID
    MICAPS_GDS_IP='xxx.xxx.xxx.xxx',       # MICAPS GDS server IP
    MICAPS_GDS_PORT='xxx',                 # MICAPS GDS port
    THREDDS_IP='xxx.xxx.xxx.xxx',          # THREDDS server IP
    THREDDS_PORT='xxx',                    # THREDDS port
    CDS_API_TOKEN='your_cds_token',        # Copernicus CDS API token
    CACHE_DIR='/path/to/cache'             # Local cache directory
)
```

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

### Generate Height, Wind, and Precipitation Analysis

Source: https://context7.com/nmcdev/metdig/llms.txt

Produces weather charts visualizing geopotential height, wind barbs, and precipitation accumulation. The atime parameter specifies the accumulation period in hours.

```python
import metdig.onestep as mdgonestep
from datetime import datetime
import matplotlib.pyplot as plt

# Generate 500hPa height + 850hPa wind + precipitation chart
result = mdgonestep.hgt_uv_rain(
    data_source='cassandra',
    data_name='ecmwf',
    init_time=datetime(2023, 7, 15, 8),
    fhour=24,
    atime=6,                               # Precipitation accumulation period (hours)
    hgt_lev=500,                           # Height contour level
    uv_lev=850,                            # Wind barbs level
    is_mask_terrain=True,
    area='华北',                           # Analysis area
    is_return_data=True,
    is_draw=True,
    output_dir='/path/to/output',
    png_name='hgt_wind_rain.png'
)
plt.show()
```

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

### Analyze Temperature Extremes

Source: https://context7.com/nmcdev/metdig/llms.txt

Visualizes 24-hour maximum and minimum temperature forecasts. Requires specific data sources like 'nwfd_scmoc'.

```python
import metdig.onestep as mdgonestep
from datetime import datetime
import matplotlib.pyplot as plt

# Generate 24-hour maximum temperature chart
result_max = mdgonestep.t2m_mx24(
    data_source='cassandra',
    data_name='nwfd_scmoc',
    init_time=datetime(2023, 7, 15, 8),
    fhour=24,
    area='全国',
    is_return_data=True,
    is_draw=True,
    output_dir='/path/to/output',
    png_name='t2m_max24.png'
)

# Generate 24-hour minimum temperature chart
result_min = mdgonestep.t2m_mn24(
    data_source='cassandra',
    data_name='nwfd_scmoc',
    init_time=datetime(2023, 7, 15, 8),
    fhour=24,
    area='全国',
    is_return_data=True,
    is_draw=True,
    output_dir='/path/to/output',
    png_name='t2m_min24.png'
)
plt.show()
```

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

### Create Skew-T Log-P Diagram

Source: https://context7.com/nmcdev/metdig/llms.txt

Generates a Skew-T Log-P diagram for thermodynamic soundings. Initializes a SkewT composition object with a title and output parameters, then allows plotting of profiles.

```python
import metdig.graphics.draw_compose as dc
import metdig.io as mdgio
from datetime import datetime
import matplotlib.pyplot as plt

# Create Skew-T composition
skew_obj = dc.skewt_compose(
    title='Model Sounding',
    description='Beijing 2023-07-15 08UTC',
    output_dir='/path/to/output',
    png_name='skewt_diagram.png'
)

# The skew object provides access to:
# - skew_obj.fig: matplotlib figure
# - skew_obj.skew: SkewT object for plotting profiles

# Save composition
result = skew_obj.save()
```

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

### Generate Time-Height Cross-Section with Divergence and Vorticity

Source: https://context7.com/nmcdev/metdig/llms.txt

Generates a time-height cross-section displaying divergence, vorticity, relative humidity, and wind. Requires specifying data source, initialization time, forecast hours, levels, and geographical points.

```python
result_dv = mdgonestep.time_div_vort_rh_uv(
    data_source='cassandra',
    data_name='ecmwf',
    init_time=datetime(2023, 7, 15, 8),
    fhours=range(0, 72, 3),
    levels=[1000, 925, 850, 700, 500, 300],
    points={'lon': [120], 'lat': [30]},
    is_draw=True
)
plt.show()
```

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

### Read Multi-Level 3D Model Grid Data

Source: https://context7.com/nmcdev/metdig/llms.txt

Retrieve 3D model data (e.g., temperature, wind components) across multiple pressure levels for a single time step. Specify the data source, model name, initialization time, forecast hour, variable name, a list of pressure levels, and geographic extent.

```python
import metdig.io as mdgio
from datetime import datetime

# Read multi-level 3D grid data for a single time
levels = [1000, 925, 850, 700, 500, 300, 200]
tmp_3d = mdgio.get_model_3D_grid(
    data_source='cassandra',
    data_name='ecmwf',
    init_time=datetime(2023, 7, 15, 8),
    fhour=24,
    var_name='tmp',                        # Temperature
    levels=levels,                         # Multiple pressure levels
    extent=[100, 130, 20, 50]
)

# Read u and v wind components
u_3d = mdgio.get_model_3D_grid(
    data_source='cassandra', data_name='ecmwf',
    init_time=datetime(2023, 7, 15, 8), fhour=24,
    var_name='u', levels=levels, extent=[100, 130, 20, 50]
)
v_3d = mdgio.get_model_3D_grid(
    data_source='cassandra', data_name='ecmwf',
    init_time=datetime(2023, 7, 15, 8), fhour=24,
    var_name='v', levels=levels, extent=[100, 130, 20, 50]
)
print(f"3D Temperature shape: {tmp_3d.shape}")
```

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

### Read Single-Level Model Grid Data

Source: https://context7.com/nmcdev/metdig/llms.txt

Retrieve single-level, single-time model grid data from specified sources like 'cassandra', 'cmadaas', 'cds', or 'thredds'. Requires model name, initialization time, forecast hour, variable name, pressure level, and geographic extent.

```python
import metdig.io as mdgio
from datetime import datetime

# Read single-level, single-time grid data
hgt_500 = mdgio.get_model_grid(
    data_source='cassandra',               # Data source: cassandra, cmadaas, cds, thredds, custom
    data_name='ecmwf',                     # Model name: ecmwf, cma_gfs, etc.
    init_time=datetime(2023, 7, 15, 8),    # Model initialization time
    fhour=24,                               # Forecast hour
    var_name='hgt',                         # Variable name: hgt, tmp, u, v, rh, etc.
    level=500,                              # Pressure level (hPa)
    extent=[60, 145, 15, 55]               # Geographic extent [lon_min, lon_max, lat_min, lat_max]
)
print(f"Data shape: {hgt_500.shape}")
print(f"Variable units: {hgt_500.attrs['var_units']}")
```

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

### Generate Synoptic Composite Analysis

Source: https://context7.com/nmcdev/metdig/llms.txt

Creates comprehensive synoptic charts combining multiple meteorological fields like height, wind, and vorticity. Set is_return_data to True to access the underlying data dictionary.

```python
import metdig.onestep as mdgonestep
from datetime import datetime
import matplotlib.pyplot as plt

# Generate synoptic composite chart with 500hPa height,
# 850hPa wind, 200hPa jet, vorticity, MSLP, and TCWV
result = mdgonestep.syn_composite(
    data_source='cassandra',
    data_name='ecmwf',
    init_time=datetime(2023, 7, 15, 8),
    fhour=24,
    hgt_lev=500,                           # Geopotential height level
    uv_lev=850,                            # Wind barbs level
    wsp_lev=200,                           # Wind speed (jet) level
    vort_lev=500,                          # Vorticity level
    is_mask_terrain=True,                  # Mask terrain-blocked regions
    area='全国',                           # Analysis area (China)
    is_return_data=True,                   # Return data for further analysis
    is_draw=True,                          # Generate visualization
    add_city=True,                         # Add city markers
    output_dir='/path/to/output',
    png_name='synoptic_composite.png'
)
plt.show()

# Access returned data
if 'data' in result:
    hgt = result['data']['hgt']
    prmsl = result['data']['prmsl']
```

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

### Read Multi-Time Model Grid Data

Source: https://context7.com/nmcdev/metdig/llms.txt

Retrieve model data for multiple forecast hours for time-series analysis. This can be single-level data (e.g., 2-meter temperature) or multi-level 3D data (e.g., temperature at various pressure levels). Specify a range or array of forecast hours.

```python
import metdig.io as mdgio
from datetime import datetime
import numpy as np

# Read single-level, multi-time grid data
fhours = np.arange(0, 73, 3)  # 0 to 72 hours, 3-hour intervals
t2m_series = mdgio.get_model_grids(
    data_source='cassandra',
    data_name='ecmwf',
    init_time=datetime(2023, 7, 15, 8),
    fhours=fhours,                         # Multiple forecast hours
    var_name='t2m',                        # 2-meter temperature
    extent=[100, 130, 20, 50]
)

# Read multi-level, multi-time 3D grid data
levels = [850, 500, 200]
tmp_4d = mdgio.get_model_3D_grids(
    data_source='cassandra',
    data_name='ecmwf',
    init_time=datetime(2023, 7, 15, 8),
    fhours=np.arange(0, 49, 6),
    var_name='tmp',
    levels=levels,
    extent=[100, 130, 20, 50]
)
print(f"4D data dimensions: {tmp_4d.dims}")
```

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

### Calculate Moisture Diagnostics

Source: https://context7.com/nmcdev/metdig/llms.txt

Computes dewpoint, specific humidity, and equivalent potential temperature from temperature and relative humidity. Requires reading temperature and relative humidity at a specific level and extent. The output includes units and ranges.

```python
import metdig.cal as mdgcal
import metdig.io as mdgio
import metdig.utl as mdgstda
from datetime import datetime

init_time = datetime(2023, 7, 15, 8)

# Read temperature and relative humidity
tmp = mdgio.get_model_grid(
    data_source='cassandra', data_name='ecmwf',
    init_time=init_time, fhour=24,
    var_name='tmp', level=850, extent=[100, 130, 20, 50]
)
rh = mdgio.get_model_grid(
    data_source='cassandra', data_name='ecmwf',
    init_time=init_time, fhour=24,
    var_name='rh', level=850, extent=[100, 130, 20, 50]
)

# Calculate dewpoint from relative humidity
td = mdgcal.dewpoint_from_relative_humidity(tmp, rh)

# Calculate specific humidity from dewpoint
pressure = mdgstda.gridstda_full_like_by_levels(td, [850])
spfh = mdgcal.specific_humidity_from_dewpoint(pressure, td)
print(f"Specific humidity units: {spfh.attrs['var_units']}")

# Calculate equivalent potential temperature
theta_e = mdgcal.equivalent_potential_temperature(pressure, tmp, td)
print(f"Theta-e range: {float(theta_e.min()):.1f} - {float(theta_e.max()):.1f} K")
```

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

### Generate Wind, Theta, and Vorticity Cross-Section

Source: https://context7.com/nmcdev/metdig/llms.txt

Generates a cross-section plot showing wind, equivalent potential temperature, and specific humidity. Requires specifying data source, initialization time, forecast hour, levels, start/end points, and averaging windows. Can optionally mask terrain and return data.

```python
result = mdgonestep.wind_theta_spfh(
    data_source='cassandra',
    data_name='ecmwf',
    init_time=datetime(2023, 7, 15, 8),
    fhour=24,
    levels=[1000, 950, 925, 900, 850, 800, 700, 600, 500, 400, 300, 200],
    st_point=[25, 110],                    # Start point [lat, lon]
    ed_point=[40, 125],                    # End point [lat, lon]
    lon_mean=1.0,                          # Longitude averaging window (degrees)
    lat_mean=1.0,                          # Latitude averaging window (degrees)
    is_mask_terrain=True,
    area='华东',
    is_return_data=True,
    is_draw=True,
    output_dir='/path/to/output',
    png_name='cross_theta_spfh.png'
)
plt.show()
```

```python
result_vort = mdgonestep.wind_theta_vort(
    data_source='cassandra',
    data_name='ecmwf',
    init_time=datetime(2023, 7, 15, 8),
    fhour=24,
    levels=[1000, 925, 850, 700, 500, 300, 200],
    st_point=[25, 110],
    ed_point=[40, 125],
    is_draw=True
)
plt.show()
```

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

### Create Custom Horizontal Map Composition

Source: https://context7.com/nmcdev/metdig/llms.txt

Composes a custom horizontal map with multiple overlays, such as height contours and wind barbs. Requires reading data for different variables and levels, then using the graphics module to add elements.

```python
import metdig.graphics.draw_compose as dc
import metdig.graphics.contourf_method as cm
import metdig.graphics.contour_method as ct
import metdig.graphics.barbs_method as bm
import metdig.io as mdgio
import metdig.cal as mdgcal
from datetime import datetime
import matplotlib.pyplot as plt

init_time = datetime(2023, 7, 15, 8)

# Read data
hgt = mdgio.get_model_grid(
    data_source='cassandra', data_name='ecmwf',
    init_time=init_time, fhour=24,
    var_name='hgt', level=500, extent=[60, 145, 15, 55]
)
u = mdgio.get_model_grid(
    data_source='cassandra', data_name='ecmwf',
    init_time=init_time, fhour=24,
    var_name='u', level=850, extent=[60, 145, 15, 55]
)
v = mdgio.get_model_grid(
    data_source='cassandra', data_name='ecmwf',
    init_time=init_time, fhour=24,
    var_name='v', level=850, extent=[60, 145, 15, 55]
)
wsp = mdgcal.other.wind_speed(u, v)

# Create horizontal composition
obj = dc.horizontal_compose(
    map_extent=(60, 145, 15, 55),
    description='500hPa Height + 850hPa Wind',
    output_dir='/path/to/output',
    png_name='custom_composite.png',
    figsize=(16, 12)
)

# Add wind speed shading
cm.wind_speed_pcolormesh(obj.ax, wsp)

# Add height contours
ct.hgt_contour(obj.ax, hgt, levels=range(5400, 5900, 40))

# Add wind barbs
bm.uv_barbs(obj.ax, u, v, regrid_shape=20)

# Save and display
obj.save()
plt.show()
```

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

### Generate Time-Height Cross-Section of RH, Wind, and Theta

Source: https://context7.com/nmcdev/metdig/llms.txt

Creates a time-height cross-section showing relative humidity, wind, and theta. Requires specifying data source, initialization time, forecast hours, levels, and geographical points. Can mask terrain and return data.

```python
import metdig.onestep as mdgonestep
from datetime import datetime
import matplotlib.pyplot as plt

# Time-height cross-section of RH, wind, and theta
result = mdgonestep.time_rh_uv_theta(
    data_source='cassandra',
    data_name='ecmwf',
    init_time=datetime(2023, 7, 15, 8),
    fhours=range(0, 72, 3),                # Forecast hours
    levels=[1000, 925, 850, 700, 500, 300, 200],
    points={'lon': [116.4], 'lat': [39.9]},  # Beijing coordinates
    mean_area=1.0,                         # Area averaging radius (degrees)
    is_mask_terrain=True,
    is_return_data=True,
    is_draw=True,
    output_dir='/path/to/output',
    png_name='time_height_rh_theta.png'
)
plt.show()
```

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

### Read Multi-Time Station Observation Data

Source: https://context7.com/nmcdev/metdig/llms.txt

Reads surface observation data for multiple time points from a specified data source. Requires specifying a list of observation times, data name, variable name, and geographic extent.

```python
obs_series = mdgio.get_obs_stations_multitime(
    data_source='cassandra',
    times=[datetime(2023, 7, 15, i) for i in range(0, 24, 3)],
    data_name='sfc_chn_hor',
    var_name='TEM',
    extent=[100, 130, 20, 50]
)
```

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

### Read Single-Time Station Observation Data

Source: https://context7.com/nmcdev/metdig/llms.txt

Reads surface observation data for a single time point from a specified data source. Requires specifying the observation time, data name, variable name, and geographic extent.

```python
obs_data = mdgio.get_obs_stations(
    data_source='cassandra',
    obs_time=datetime(2023, 7, 15, 8),
    data_name='sfc_chn_hor',               # Surface observation data type
    var_name='TEM',                        # Temperature variable
    extent=[100, 130, 20, 50]
)
```

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

### Set Log Level for Debugging

Source: https://context7.com/nmcdev/metdig/llms.txt

Configures the logging level for the MetDig library to 'debug' to enable detailed output for troubleshooting data retrieval and computation issues. This setting affects subsequent operations.

```python
import metdig

# Set log level for debugging
metdig.set_loglevel('debug')    # Options: notset, debug, info, warning, error, critical

# Now all metdig operations will output detailed logs
import metdig.io as mdgio
from datetime import datetime

data = mdgio.get_model_grid(
    data_source='cassandra',
    data_name='ecmwf',
    init_time=datetime(2023, 7, 15, 8),
    fhour=24,
    var_name='hgt',
    level=500
)
```

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

### Calculate Potential Vorticity

Source: https://context7.com/nmcdev/metdig/llms.txt

Computes baroclinic potential vorticity (PVU) using temperature, pressure, and wind components. Requires reading temperature, U, and V wind fields at specified levels and extent. The output includes units.

```python
import metdig.cal as mdgcal
import metdig.io as mdgio
import metdig.utl as mdgstda
from datetime import datetime

levels = [700, 500, 300, 200]
init_time = datetime(2023, 7, 15, 8)

# Read required fields
tmp = mdgio.get_model_3D_grid(
    data_source='cassandra', data_name='ecmwf',
    init_time=init_time, fhour=24,
    var_name='tmp', levels=levels, extent=[100, 130, 20, 50]
)
u = mdgio.get_model_3D_grid(
    data_source='cassandra', data_name='ecmwf',
    init_time=init_time, fhour=24,
    var_name='u', levels=levels, extent=[100, 130, 20, 50]
)
v = mdgio.get_model_3D_grid(
    data_source='cassandra', data_name='ecmwf',
    init_time=init_time, fhour=24,
    var_name='v', levels=levels, extent=[100, 130, 20, 50]
)

# Create pressure field and calculate potential temperature
pressure = mdgstda.gridstda_full_like_by_levels(tmp, levels)
theta = mdgcal.potential_temperature(pressure, tmp)

# Calculate baroclinic potential vorticity (PVU)
pv = mdgcal.potential_vorticity_baroclinic(theta, pressure, u, v)
print(f"PV units: {pv.attrs['var_units']}")
```

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

### Calculate Vorticity and Divergence

Source: https://context7.com/nmcdev/metdig/llms.txt

Computes vorticity and divergence from U and V wind components. Ensure U and V are read with appropriate levels and extent. The output includes units and maximum values.

```python
import metdig.cal as mdgcal
import metdig.io as mdgio
from datetime import datetime

# Read wind components
levels = [850, 500]
u = mdgio.get_model_3D_grid(
    data_source='cassandra', data_name='ecmwf',
    init_time=datetime(2023, 7, 15, 8), fhour=24,
    var_name='u', levels=levels, extent=[100, 130, 20, 50]
)
v = mdgio.get_model_3D_grid(
    data_source='cassandra', data_name='ecmwf',
    init_time=datetime(2023, 7, 15, 8), fhour=24,
    var_name='v', levels=levels, extent=[100, 130, 20, 50]
)

# Calculate vorticity (1/s)
vort = mdgcal.vorticity(u, v)
print(f"Vorticity units: {vort.attrs['var_units']}")

# Calculate absolute vorticity (includes planetary vorticity)
absv = mdgcal.absolute_vorticity(u, v)

# Calculate divergence (1/s)
div = mdgcal.divergence(u, v)
print(f"Divergence max: {float(div.max()):.2e}")
```

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

### Calculate Advection

Source: https://context7.com/nmcdev/metdig/llms.txt

Computes horizontal advection of a scalar variable using wind components. Requires reading the scalar variable (e.g., temperature) and U, V wind components at a specific level and extent. The output includes units.

```python
import metdig.cal as mdgcal
import metdig.io as mdgio
from datetime import datetime

init_time = datetime(2023, 7, 15, 8)
level = 850

# Read temperature and wind
tmp = mdgio.get_model_grid(
    data_source='cassandra', data_name='ecmwf',
    init_time=init_time, fhour=24,
    var_name='tmp', level=level, extent=[100, 130, 20, 50]
)
u = mdgio.get_model_grid(
    data_source='cassandra', data_name='ecmwf',
    init_time=init_time, fhour=24,
    var_name='u', level=level, extent=[100, 130, 20, 50]
)
v = mdgio.get_model_grid(
    data_source='cassandra', data_name='ecmwf',
    init_time=init_time, fhour=24,
    var_name='v', level=level, extent=[100, 130, 20, 50]
)

# Calculate temperature advection
tmp_adv = mdgcal.var_advect(tmp, u, v)
print(f"Temperature advection units: {tmp_adv.attrs['var_units']}")
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.