Parse TSWCs from CSV

This example shows how to parse a Time Series Wind Climate (TSWC) from a .csv file.

Prepare Sample Data

First, we need to prepare the sample data. This is done by creating a DataFrame with the wind speed and direction data.

import tempfile
from pathlib import Path

import numpy as np
import pandas as pd
import windkit as wk


size = 8760  # 1 year of hourly data

df = pd.DataFrame(
    {
        "wspd_50": np.random.uniform(5, 15, size=size),  # Wind speed in m/s
        "wdir_50": np.random.uniform(0, 360, size=size),  # Wind direction in degrees
    },
    index=pd.date_range("2023-01-01", periods=size, freq="h"),
)
df.index.name = "time"  # Set the index name to 'time'
print(df.head())

                       wspd_50     wdir_50
time
2023-01-01 00:00:00  12.548563   65.613975
2023-01-01 01:00:00  10.577625    3.782416
2023-01-01 02:00:00   9.628012   96.710197
2023-01-01 03:00:00  10.617875  247.375524
2023-01-01 04:00:00  11.593405   47.362309

Save to .csv

We save the DataFrame to a .csv file in a temporary directory.

tmpdir = tempfile.TemporaryDirectory()
csv_path = Path(tmpdir.name) / "sample_tswc.csv"
df.to_csv(csv_path, index=True)

Parse .csv to TSWC

Now, we can parse the CSV file to create a Time Series Wind Climate (TSWC) object using the windkit.read_tswc() function. We have to specify the spatial coordinate system (CRS), coordinates for the west-east and south-north axes, the time column, and the height to columns mapping. In this case, we assume that the wind speed and direction data are at a height of 50 meters, and we map them accordingly.

tswc = wk.read_tswc(
    csv_path,
    west_east=0.0,
    south_north=0.0,
    crs=4326,
    time_col=0,
    height_to_columns={50: ["wspd_50", "wdir_50"]},
)

print(tswc)

<xarray.Dataset> Size: 210kB
Dimensions:         (time: 8760, height: 1, stacked_point: 1)
Coordinates:
  * time            (time) datetime64[us] 70kB 2023-01-01 ... 2023-12-31T23:0...
  * height          (height) int64 8B 50
    west_east       (stacked_point) float64 8B 0.0
    south_north     (stacked_point) float64 8B 0.0
    crs             int8 1B 0
Dimensions without coordinates: stacked_point
Data variables:
    wind_speed      (time, height, stacked_point) float64 70kB 12.55 ... 7.781
    wind_direction  (time, height, stacked_point) float64 70kB 65.61 ... 129.1
Attributes:
    history:          2026-09-29T09:45:32+00:00:\twindkit==2.2.0\t    return ...
    Conventions:      CF-1.8
    Package name:     windkit
    Package version:  2.2.0
    Creation date:    2026-09-29T09:45:32+00:00
    Object type:      Time Series Wind Climate

Alternatively, we can also use the windkit.tswc_from_dataframe() function to parse a DataFrame directly into a TSWC object. This is useful if you already have the data in a DataFrame format and do not need to read it from a file.

tswc_from_df = wk.tswc_from_dataframe(
    df,
    west_east=0.0,
    south_north=0.0,
    crs=4326,
    height_to_columns={50: ["wspd_50", "wdir_50"]},
)
print(tswc_from_df)

tmpdir.cleanup()

<xarray.Dataset> Size: 210kB
Dimensions:         (time: 8760, height: 1, stacked_point: 1)
Coordinates:
  * time            (time) datetime64[us] 70kB 2023-01-01 ... 2023-12-31T23:0...
  * height          (height) int64 8B 50
    west_east       (stacked_point) float64 8B 0.0
    south_north     (stacked_point) float64 8B 0.0
    crs             int8 1B 0
Dimensions without coordinates: stacked_point
Data variables:
    wind_speed      (time, height, stacked_point) float64 70kB 12.55 ... 7.781
    wind_direction  (time, height, stacked_point) float64 70kB 65.61 ... 129.1
Attributes:
    history:          2026-09-29T09:45:32+00:00:\twindkit==2.2.0\twk.tswc_fro...
    Conventions:      CF-1.8
    Package name:     windkit
    Package version:  2.2.0
    Creation date:    2026-09-29T09:45:32+00:00
    Object type:      Time Series Wind Climate