Skip to content

NGI Data Source

C pynasonde.vipir.ngi.source — Trace, Dataset, DataSource — data model for NGI ionogram traces.

pynasonde.vipir.ngi.source

Data ingestion helpers for VIPIR NGI ionogram datasets.

Provides structures to represent scaled traces (Trace), the full NGI dataset (Dataset), and a DataSource orchestrator that loads raw NGI files, extracts plots, and writes scaled diagnostics.

Trace dataclass

Container for scaled trace parameters extracted from NGI files.

Attributes:

Name Type Description
traces pd.DataFrame

Data frame of individual trace samples.
trace_params pd.DataFrame

Data frame describing fitted trace parameters.
Source code in pynasonde/vipir/ngi/source.py 
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
@dataclass
class Trace:
    """Container for scaled trace parameters extracted from NGI files.

    Attributes:
        traces: Data frame of individual trace samples.
        trace_params: Data frame describing fitted trace parameters.
    """

    traces: pd.DataFrame = None
    trace_params: pd.DataFrame = None

    @staticmethod
    def load_saved_scaled_parameters(
        folder: str,
        extension="*.nc",
        mode: str = "O",
        local_tz: str = None,
        lat: float = 37.8815,
        long: float = -75.4374,
    ):
        """Load previously scaled trace parameters from NetCDF files.

        Args:
            folder: Directory containing saved scaling products.
            extension: File extension glob (default ``*.nc``).
            mode: Wave mode prefix (e.g., ``'O'`` or ``'X'``).
            local_tz: Optional local timezone name. If omitted the value is
                inferred from latitude/longitude.
            lat: Station latitude in degrees.
            long: Station longitude in degrees.

        Returns:
            Pandas DataFrame with time, local time, solar zenith angle, and
            fitted frequency/height pairs.
        """
        files = glob.glob(os.path.join(folder, mode + extension))
        files.sort()
        traces = {
            "sza": [],
            "time": [],
            "fs": [],
            "hs": [],
            "local_time": [],
        }
        LTC = TimeZoneConversion(local_tz, lat, long)
        for file in files:
            d = xr.open_dataset(file)
            L = len(d.hs.values)
            traces["time"].extend(d.time.values.tolist() * L)
            traces["local_time"].extend(
                LTC.utc_to_local_time(pd.to_datetime(d.time.values).tolist()) * L
            )
            traces["sza"].extend(d.sza.values.tolist() * L)
            traces["hs"].extend(d.hs.values.tolist())
            traces["fs"].extend(d.fs.values.tolist())
        traces = pd.DataFrame.from_dict(traces)
        traces.time = pd.to_datetime(traces.time)
        return traces

load_saved_scaled_parameters(folder, extension='*.nc', mode='O', local_tz=None, lat=37.8815, long=-75.4374) staticmethod

Load previously scaled trace parameters from NetCDF files.

Parameters:

Name Type Description Default
folder str

Directory containing saved scaling products.

 required
extension

File extension glob (default *.nc).

 '*.nc'
mode str

Wave mode prefix (e.g., 'O' or 'X').

 'O'
local_tz str

Optional local timezone name. If omitted the value is inferred from latitude/longitude.

 None
lat float

Station latitude in degrees.

 37.8815
long float

Station longitude in degrees.

 -75.4374

Returns:

Type Description

Pandas DataFrame with time, local time, solar zenith angle, and

fitted frequency/height pairs.
Source code in pynasonde/vipir/ngi/source.py 
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
@staticmethod
def load_saved_scaled_parameters(
    folder: str,
    extension="*.nc",
    mode: str = "O",
    local_tz: str = None,
    lat: float = 37.8815,
    long: float = -75.4374,
):
    """Load previously scaled trace parameters from NetCDF files.

    Args:
        folder: Directory containing saved scaling products.
        extension: File extension glob (default ``*.nc``).
        mode: Wave mode prefix (e.g., ``'O'`` or ``'X'``).
        local_tz: Optional local timezone name. If omitted the value is
            inferred from latitude/longitude.
        lat: Station latitude in degrees.
        long: Station longitude in degrees.

    Returns:
        Pandas DataFrame with time, local time, solar zenith angle, and
        fitted frequency/height pairs.
    """
    files = glob.glob(os.path.join(folder, mode + extension))
    files.sort()
    traces = {
        "sza": [],
        "time": [],
        "fs": [],
        "hs": [],
        "local_time": [],
    }
    LTC = TimeZoneConversion(local_tz, lat, long)
    for file in files:
        d = xr.open_dataset(file)
        L = len(d.hs.values)
        traces["time"].extend(d.time.values.tolist() * L)
        traces["local_time"].extend(
            LTC.utc_to_local_time(pd.to_datetime(d.time.values).tolist()) * L
        )
        traces["sza"].extend(d.sza.values.tolist() * L)
        traces["hs"].extend(d.hs.values.tolist())
        traces["fs"].extend(d.fs.values.tolist())
    traces = pd.DataFrame.from_dict(traces)
    traces.time = pd.to_datetime(traces.time)
    return traces

Dataset dataclass

Representation of a single NGI dataset and its derived traces.

Source code in pynasonde/vipir/ngi/source.py 
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
@dataclass
class Dataset:
    """Representation of a single NGI dataset and its derived traces."""

    URSI: str = ""
    StationName: str = ""
    year: int = 1970  # UTC
    daynumber: int = 1  # UTC
    month: int = 1  # UTC
    day: int = 1  # UTC
    hour: int = 0  # UTC
    minute: int = 0  # UTC
    second: int = 0  # UTC
    epoch: np.datetime64 = 0  # UTC
    latitude: float = 0.0  # degree_north
    longitude: float = 0.0  # degree_east
    altitude: float = 0.0  # meter
    MagLat: float = 0.0  # degree_east
    MagLon: float = 0.0  # degree_east
    MagDip: float = 0.0  # degree
    GyroFreq: float = 0.0  # Station GyroFrequency at 300 km altitude, MHz
    range_gate_offset: float = 0.0  # microsecond
    gate_count: float = 0.0  # counts
    gate_start: float = 0.0  # microsecond
    gate_end: float = 0.0  # microsecond
    gate_step: float = 0.0  # microsecond
    Range0: float = 0.0  # kilometer
    freq_start: float = 0.0  # lower frequency, kilohertz
    freq_end: float = 0.0  # upper frequency, kilohertz
    tune_type: int = 0  # 1=log 2=linear 3=table
    freq_count: int = 0  # count
    linear_step: float = 0.0  # kilohertz
    log_step: float = 0.0  # logarithmic tuning step, percent
    Range: np.array = None  # kilometer
    Frequency: np.array = None  # kilohertz
    Time: np.array = None  # Nominal Observation Time, UT / second
    TxDrive: np.array = None  # decibel
    NumAve: np.array = None  # count
    SCT_version: float = 1.2  #
    SCT: int = 0
    PREFACE: int = 0
    Has_total_power: int = 0  # flag
    total_power: np.array = None  # decibel
    total_noise: np.array = None  # decibel
    Has_O_mode_power: int = 0  # flag
    O_mode_power: np.array = None  # decibel, Shape(Frequency, Range)
    O_mode_noise: np.array = None  # decibel, Shape(Frequency, )
    Has_X_mode_power: int = 0  # flag
    X_mode_power: np.array = None  # decibel, Shape(Frequency, Range)
    X_mode_noise: np.array = None  # decibel, Shape(Frequency, )
    Has_Doppler: int = 0  # flag
    Has_VLoS: int = 0  # flag
    Has_SPGR: int = 0  # flag
    Has_Zenith: int = 0  # flag
    Has_Azimuth: int = 0  # flag
    Has_Coherence: int = 0  # flag

    def __initialize__(self, ds, unicode: str = "latin-1"):
        """Populate the dataclass fields using an xarray dataset.

        Args:
            ds: `xarray.Dataset` produced from an NGI NetCDF file.
            unicode: Encoding used for legacy byte-string attributes.

        Returns:
            Dataset: The populated instance (also stored as ``self``).
        """
        key_map = {
            "Has_O_mode_power": "Has_O-mode_power",
            "O_mode_power": "O-mode_power",
            "O_mode_noise": "O-mode_noise",
            "Has_X_mode_power": "Has_X-mode_power",
            "X_mode_power": "X-mode_power",
            "X_mode_noise": "X-mode_noise",
        }
        for attr in self.__dict__.keys():
            if attr in list(key_map.keys()):
                setattr(self, attr, np.array(ds[key_map[attr]].values))
            elif (type(ds[attr].values) == np.ndarray) and (
                (ds[attr].values.dtype == "|S8") or (ds[attr].values.dtype == "|S64")
            ):
                setattr(self, attr, ds[attr].values.astype(str).tolist())
            else:
                if len(ds[attr].values.shape) == 0:
                    setattr(self, attr, ds[attr].values.tolist())
                else:
                    setattr(self, attr, np.array(ds[attr].values))
        self.StationName = "".join([u.decode("latin-1") for u in self.StationName])
        self.URSI = "".join([u.decode("latin-1") for u in self.URSI])
        self.time = dt.datetime(
            self.year,
            self.month,
            self.day,
            self.hour,
            self.minute,
            self.second,
        )
        self.sza = 90.0 - get_altitude(
            self.latitude,
            self.longitude,
            self.time.replace(
                tzinfo=dt.timezone.utc,
            ),
        )
        logger.info(f"Local zenith angle: {self.sza}")
        return self

    def set_traces(self, traces: dict, trace_params: dict) -> None:
        """Attach fitted trace data produced by the autoscaler.

        Args:
            traces: Mapping of cluster labels to extracted trace samples.
            trace_params: Mapping of cluster labels to metadata dictionaries.
        """
        self.trace = Trace()
        self.trace.traces = pd.concat([traces[k] for k in list(traces.keys())])
        self.trace.trace_params = pd.DataFrame.from_records(
            [trace_params[k] for k in list(trace_params.keys())]
        )
        return

    def get_n_traces(self):
        """Return the number of trace parameter sets currently stored.

        Returns:
            int: Number of available traces (zero when none were set).
        """
        _n = 0
        if hasattr(self, "trace"):
            _n = len(self.trace.trace_params)
        return _n

__initialize__(ds, unicode='latin-1')

Populate the dataclass fields using an xarray dataset.

Parameters:

Name Type Description Default
ds

xarray.Dataset produced from an NGI NetCDF file.

 required
unicode str

Encoding used for legacy byte-string attributes.

 'latin-1'

Returns:

Name Type Description
Dataset

The populated instance (also stored as self).
Source code in pynasonde/vipir/ngi/source.py 
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
def __initialize__(self, ds, unicode: str = "latin-1"):
    """Populate the dataclass fields using an xarray dataset.

    Args:
        ds: `xarray.Dataset` produced from an NGI NetCDF file.
        unicode: Encoding used for legacy byte-string attributes.

    Returns:
        Dataset: The populated instance (also stored as ``self``).
    """
    key_map = {
        "Has_O_mode_power": "Has_O-mode_power",
        "O_mode_power": "O-mode_power",
        "O_mode_noise": "O-mode_noise",
        "Has_X_mode_power": "Has_X-mode_power",
        "X_mode_power": "X-mode_power",
        "X_mode_noise": "X-mode_noise",
    }
    for attr in self.__dict__.keys():
        if attr in list(key_map.keys()):
            setattr(self, attr, np.array(ds[key_map[attr]].values))
        elif (type(ds[attr].values) == np.ndarray) and (
            (ds[attr].values.dtype == "|S8") or (ds[attr].values.dtype == "|S64")
        ):
            setattr(self, attr, ds[attr].values.astype(str).tolist())
        else:
            if len(ds[attr].values.shape) == 0:
                setattr(self, attr, ds[attr].values.tolist())
            else:
                setattr(self, attr, np.array(ds[attr].values))
    self.StationName = "".join([u.decode("latin-1") for u in self.StationName])
    self.URSI = "".join([u.decode("latin-1") for u in self.URSI])
    self.time = dt.datetime(
        self.year,
        self.month,
        self.day,
        self.hour,
        self.minute,
        self.second,
    )
    self.sza = 90.0 - get_altitude(
        self.latitude,
        self.longitude,
        self.time.replace(
            tzinfo=dt.timezone.utc,
        ),
    )
    logger.info(f"Local zenith angle: {self.sza}")
    return self

set_traces(traces, trace_params)

Attach fitted trace data produced by the autoscaler.

Parameters:

Name Type Description Default
traces dict

Mapping of cluster labels to extracted trace samples.

 required
trace_params dict

Mapping of cluster labels to metadata dictionaries.

 required
Source code in pynasonde/vipir/ngi/source.py 
196
197
198
199
200
201
202
203
204
205
206
207
208
def set_traces(self, traces: dict, trace_params: dict) -> None:
    """Attach fitted trace data produced by the autoscaler.

    Args:
        traces: Mapping of cluster labels to extracted trace samples.
        trace_params: Mapping of cluster labels to metadata dictionaries.
    """
    self.trace = Trace()
    self.trace.traces = pd.concat([traces[k] for k in list(traces.keys())])
    self.trace.trace_params = pd.DataFrame.from_records(
        [trace_params[k] for k in list(trace_params.keys())]
    )
    return

get_n_traces()

Return the number of trace parameter sets currently stored.

Returns:

Name Type Description
int

Number of available traces (zero when none were set).
Source code in pynasonde/vipir/ngi/source.py 
210
211
212
213
214
215
216
217
218
219
def get_n_traces(self):
    """Return the number of trace parameter sets currently stored.

    Returns:
        int: Number of available traces (zero when none were set).
    """
    _n = 0
    if hasattr(self, "trace"):
        _n = len(self.trace.trace_params)
    return _n

DataSource

Bases: object

Manage discovery and loading of NGI ionogram files.

Source code in pynasonde/vipir/ngi/source.py 
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
class DataSource(object):
    """Manage discovery and loading of NGI ionogram files."""

    def __init__(
        self,
        source_folder: str = "./tmp/",
        file_ext: str = "*.ngi.bz2",
        file_names: List[str] = [],
        needs_decompression: bool = False,
    ):
        """Configure the datasource to point at a folder or explicit files.

        Args:
            source_folder: Directory containing NGI files.
            file_ext: Filename glob to match ionogram products.
            file_names: Explicit list of filenames (optional).
            needs_decompression: Force decompression even if extension differs.
        """
        self.source_folder = source_folder
        self.file_ext = file_ext
        self.file_names = file_names
        self.needs_decompression = (
            True if (".bz2" in file_ext) or needs_decompression else False
        )
        # Load full path of the files by file_names or from free space search
        self.file_paths = (
            [os.path.join(source_folder, f) for f in file_names]
            if file_names
            else glob.glob(os.path.join(source_folder, file_ext.lower()))
            + glob.glob(os.path.join(source_folder, file_ext.upper()))
        )
        self.file_paths.sort()
        logger.info(f"Total number of files {len(self.file_paths)}")
        logger.info(f"Needs decompression {self.needs_decompression}")
        return

    def load_data_sets(self, load_start=0, load_end=-1, n_jobs=-1):
        """Populate `self.datasets` with NGI records from disk.

        Args:
            load_start: Start index when slicing the discovered file list.
            load_end: End index (non-inclusive) for the slice; ``-1`` means all.
            n_jobs: Parallelism level passed to `joblib.Parallel`.
        """

        self.datasets = []
        compress = lambda fc, fd: shutil.copyfileobj(
            open(fd, "rb"), bz2.BZ2File(fc, "wb")
        )
        decompress = lambda fc, fd: shutil.copyfileobj(
            bz2.BZ2File(fc, "rb"), open(fd, "wb")
        )
        check_bad_file = lambda f: (
            True if os.path.getsize(f) / (1024 * 1024) >= 5.0 else False
        )

        def process_file(f):
            if not check_bad_file(f):
                return None
            logger.info(f"Load file: {f}")
            needs_decompression = self.needs_decompression
            orig_f = f
            if needs_decompression:
                decompress(f, f.replace(".bz2", ""))
                os.remove(f)
                f = f.replace(".bz2", "")
            ds = Dataset().__initialize__(xr.load_dataset(f, engine="netcdf4"))
            if needs_decompression:
                compress(f + ".bz2", f)
                os.remove(f)
            return ds

        files = self.file_paths[load_start:load_end]
        results = Parallel(n_jobs=n_jobs)(delayed(process_file)(f) for f in files)
        self.datasets = [ds for ds in results if ds is not None]
        return

    def extract_ionograms(self, folder: str = "tmp/", mode: str = "O") -> None:
        """Render quick-look ionograms for each dataset to PNG files.

        Args:
            folder: Destination directory for PNG images.
            mode: Wave mode to visualize (``'O'`` or ``'X'``).
        """
        os.makedirs(folder, exist_ok=True)
        for ds in self.datasets:
            time = dt.datetime(ds.year, ds.month, ds.day, ds.hour, ds.minute, ds.second)
            i = Ionogram()
            i.add_ionogram(
                ds.Frequency, ds.Range, getattr(ds, f"{mode}_mode_power"), mode=mode
            )
            i.save(
                os.path.join(folder, f"{ds.URSI}_{time.strftime('%Y%m%d%H%M%S')}.png")
            )
            i.close()
        return

    def extract_FTI_RTI(
        self,
        rlim: List[float] = [50, 800],
        flim: List[float] = [],
        mode: str = "O",
        noise_scale: float = 1.0,
    ) -> pd.DataFrame:
        """Extract frequency–time ionogram summary points for each dataset.

        Args:
            rlim: Height limits (km) used when scanning for maxima.
            flim: Optional frequency limits (MHz) for filtering results.
            mode: Ionogram mode to process.
            noise_scale: Multiplier applied to the noise threshold.

        Returns:
            DataFrame of RTI points with timestamp, frequency, power, and range.
        """
        logger.info(f"Extract FTI/RTI, based on {rlim}km")
        rti = []
        for ds in self.datasets:
            time = dt.datetime(ds.year, ds.month, ds.day, ds.hour, ds.minute, ds.second)
            logger.info(f"Time: {time}")
            f_max = np.nan
            for i, r in enumerate(ds.Range):
                if (r < rlim[0]) or (r > rlim[1]):
                    continue
                powr = np.array(getattr(ds, f"{mode}_mode_power")[:, i])
                powr[powr < getattr(ds, f"{mode}_mode_noise") * noise_scale] = (
                    np.nan
                )  # remove noise
                f_max = ds.Frequency[np.nanargmax(powr)]
                rti.append(
                    dict(
                        time=time,
                        frequency=f_max / 1e3,  # to MHz
                        power=np.nanmax(powr),  # in dB
                        range=r,
                        noise=getattr(ds, f"{mode}_mode_noise")[
                            np.nanargmax(powr)
                        ],  # in dB
                    )
                )
        rti = pd.DataFrame.from_records(rti)
        if len(flim) == 2:
            rti = rti[(rti.frequency >= flim[0]) & (rti.frequency <= flim[1])]
        return rti

    def extract_Power_RTI(
        self,
        folder: str = "tmp/",
        rlim: List[float] = [50, 800],
        flim: List[float] = [3.95, 4.05],
        mode: str = "O",
        fname: str = None,
        xlabel: str = "Time, UT",
        ylabel: str = "Virtual Height, km",
        xlim: List[dt.datetime] = None,
        add_cbar: bool = True,
        cbar_label: str = "{}-mode Power, dB",
        cmap: str = "Spectral",
        xtick_locator: mdates.HourLocator = mdates.HourLocator(interval=4),
        prange: List[float] = [5, 70],
        noise_scale: float = 1.2,
        date_format: str = r"$%H^{%M}$",
        del_ticks: bool = False,
        xdate_lims: List[dt.datetime] = None,
    ) -> pd.DataFrame:
        """Generate power RTI plots and return the underlying dataframe.

        Args:
            folder: Directory where the PNG plot will be written.
            rlim: Range limits (km) kept in the dataset.
            flim: Frequency limits (MHz) used to narrow the data.
            mode: Ionogram mode (``'O'`` or ``'X'``).
            fname: Optional filename for the saved figure.
            xlabel: X-axis label text.
            ylabel: Y-axis label text.
            xlim: Datetime limits applied to the plot.
            add_cbar: Whether to include a colorbar.
            cbar_label: Format string for the colorbar label.
            cmap: Matplotlib colormap name.
            xtick_locator: Locator controlling x-axis tick frequency.
            prange: Min/max dB levels displayed.
            noise_scale: Multiplier applied to the noise floor when masking.
            date_format: Major tick label format.
            del_ticks: Remove axis ticks before plotting (useful for grids).
            xdate_lims: Override for x-axis limits.

        Returns:
            DataFrame containing the filtered RTI samples.
        """
        logger.info(f"Extract Power/RTI, based on {flim}MHz {rlim}km")
        rti = pd.DataFrame()
        for ds in self.datasets:
            time = dt.datetime(ds.year, ds.month, ds.day, ds.hour, ds.minute, ds.second)
            logger.info(f"Time: {time}")
            frequency, range = np.meshgrid(ds.Frequency, ds.Range, indexing="ij")
            noise, _ = np.meshgrid(
                getattr(ds, f"{mode}_mode_noise"), ds.Range, indexing="ij"
            )
            o = pd.DataFrame()
            (
                o["frequency"],
                o["range"],
                o[f"{mode}_mode_power"],
                o[f"{mode}_mode_noise"],
            ) = (
                frequency.ravel() / 1e3,  # to MHz
                range.ravel(),  # in km
                getattr(ds, f"{mode}_mode_power").ravel(),  # in dB
                noise.ravel(),  # in dB
            )
            o["time"] = time
            if (len(rlim) == 2) and (len(flim) == 2):
                o = o[
                    (o.range >= rlim[0])
                    & (o.range <= rlim[1])
                    & (o.frequency >= flim[0])
                    & (o.frequency <= flim[1])
                ]
            rti = pd.concat([rti, o])

        if fname is None:
            fname = f"{ds.URSI}_{rti.time.min().strftime('%Y%m%d.%H%M-')}{rti.time.max().strftime('%H%M')}_{mode}-mode.png"
        if xdate_lims is None:
            fig_title = f"""{ds.URSI}/{rti.time.min().strftime('%H%M-')}{rti.time.max().strftime('%H%M')} UT, {rti.time.max().strftime('%d %b %Y')}"""
        else:
            fig_title = f"""{ds.URSI}/{xdate_lims[0].strftime('%H%M-')}{xdate_lims[1].strftime('%H%M')} UT, {xdate_lims[1].strftime('%d %b %Y')}"""
        fig_title += (
            r"/ $f_0\sim$[" + "%.2f" % flim[0] + "-" + "%.2f" % flim[1] + "] MHz"
        )
        i = Ionogram(fig_title=fig_title, nrows=1, ncols=1)
        i.add_interval_plots(
            rti,
            mode=mode,
            xlabel=xlabel,
            ylabel=ylabel,
            xlim=xlim,
            add_cbar=add_cbar,
            cbar_label=cbar_label.format(mode),
            cmap=cmap,
            prange=prange,
            noise_scale=noise_scale,
            del_ticks=del_ticks,
            date_format=date_format,
            ylim=rlim,
            xtick_locator=xtick_locator,
            xdate_lims=xdate_lims,
        )
        i.save(os.path.join(folder, fname))
        i.close()
        return rti

    def save_scaled_dataset(self, params=["hmF2", "foF2"]):
        """Persist selected scaled parameters to a CSV file.

        Args:
            params: Sequence of attribute names to extract per dataset.
        """
        records = []
        for ds in self.datasets:
            rec = dict()
            for p in params:
                rec[p] = getattr(ds, p)
            records.append(rec)
        records = pd.DataFrame.from_records(records)
        fname = os.path.join(self.source_folder, "scaled.csv")
        records.to_csv(fname, float_format="%g", header=True, index=False)
        return

    def save_scaled_parameters(self, attr: dict = dict(), mode: str = "O"):
        """Write trace parameters into per-epoch NetCDF files.

        Args:
            attr: Extra dataset-level attributes to store in each NetCDF file.
            mode: Mode suffix applied to the output filenames.
        """
        folder = os.path.join(self.source_folder, "scaled")
        os.makedirs(folder, exist_ok=True)
        for ds in self.datasets:
            fname = os.path.join(
                folder, f"{mode}_mode_sp_{ds.time.strftime('%Y%m%d%H%M%S')}.nc"
            )
            logger.info(f"Saved to {fname}")
            max_n_trace = ds.get_n_traces()
            fs, hs = (
                np.zeros(max_n_trace) * np.nan,
                np.zeros(max_n_trace) * np.nan,
            )
            if ds.get_n_traces():
                fs, hs = (
                    ds.trace.trace_params.fs,
                    ds.trace.trace_params.hs,
                )
            data = {
                "fs": (("max_n_trace"), fs),
                "hs": (("max_n_trace"), hs),
                "sza": (("time",), [ds.sza]),
            }
            coords = dict(time=[ds.time], max_n_trace=np.arange(max_n_trace))
            d = xr.Dataset(data, coords=coords, attrs=attr)
            d.fs.attrs["units"], d.hs.attrs["units"] = "MHz", "km"
            d.fs.attrs["description"], d.hs.attrs["description"] = (
                "Plasma Frequencies",
                "Virtual Height/Reflection Height",
            )
            d.to_netcdf(fname)
            del d
        return

__init__(source_folder='./tmp/', file_ext='*.ngi.bz2', file_names=[], needs_decompression=False)

Configure the datasource to point at a folder or explicit files.

Parameters:

Name Type Description Default
source_folder str

Directory containing NGI files.

 './tmp/'
file_ext str

Filename glob to match ionogram products.

 '*.ngi.bz2'
file_names List[str]

Explicit list of filenames (optional).

 []
needs_decompression bool

Force decompression even if extension differs.

 False
Source code in pynasonde/vipir/ngi/source.py 
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
def __init__(
    self,
    source_folder: str = "./tmp/",
    file_ext: str = "*.ngi.bz2",
    file_names: List[str] = [],
    needs_decompression: bool = False,
):
    """Configure the datasource to point at a folder or explicit files.

    Args:
        source_folder: Directory containing NGI files.
        file_ext: Filename glob to match ionogram products.
        file_names: Explicit list of filenames (optional).
        needs_decompression: Force decompression even if extension differs.
    """
    self.source_folder = source_folder
    self.file_ext = file_ext
    self.file_names = file_names
    self.needs_decompression = (
        True if (".bz2" in file_ext) or needs_decompression else False
    )
    # Load full path of the files by file_names or from free space search
    self.file_paths = (
        [os.path.join(source_folder, f) for f in file_names]
        if file_names
        else glob.glob(os.path.join(source_folder, file_ext.lower()))
        + glob.glob(os.path.join(source_folder, file_ext.upper()))
    )
    self.file_paths.sort()
    logger.info(f"Total number of files {len(self.file_paths)}")
    logger.info(f"Needs decompression {self.needs_decompression}")
    return

load_data_sets(load_start=0, load_end=-1, n_jobs=-1)

Populate self.datasets with NGI records from disk.

Parameters:

Name Type Description Default
load_start

Start index when slicing the discovered file list.

 0
load_end

End index (non-inclusive) for the slice; -1 means all.

 -1
n_jobs

Parallelism level passed to joblib.Parallel.

 -1
Source code in pynasonde/vipir/ngi/source.py 
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
def load_data_sets(self, load_start=0, load_end=-1, n_jobs=-1):
    """Populate `self.datasets` with NGI records from disk.

    Args:
        load_start: Start index when slicing the discovered file list.
        load_end: End index (non-inclusive) for the slice; ``-1`` means all.
        n_jobs: Parallelism level passed to `joblib.Parallel`.
    """

    self.datasets = []
    compress = lambda fc, fd: shutil.copyfileobj(
        open(fd, "rb"), bz2.BZ2File(fc, "wb")
    )
    decompress = lambda fc, fd: shutil.copyfileobj(
        bz2.BZ2File(fc, "rb"), open(fd, "wb")
    )
    check_bad_file = lambda f: (
        True if os.path.getsize(f) / (1024 * 1024) >= 5.0 else False
    )

    def process_file(f):
        if not check_bad_file(f):
            return None
        logger.info(f"Load file: {f}")
        needs_decompression = self.needs_decompression
        orig_f = f
        if needs_decompression:
            decompress(f, f.replace(".bz2", ""))
            os.remove(f)
            f = f.replace(".bz2", "")
        ds = Dataset().__initialize__(xr.load_dataset(f, engine="netcdf4"))
        if needs_decompression:
            compress(f + ".bz2", f)
            os.remove(f)
        return ds

    files = self.file_paths[load_start:load_end]
    results = Parallel(n_jobs=n_jobs)(delayed(process_file)(f) for f in files)
    self.datasets = [ds for ds in results if ds is not None]
    return

extract_ionograms(folder='tmp/', mode='O')

Render quick-look ionograms for each dataset to PNG files.

Parameters:

Name Type Description Default
folder str

Destination directory for PNG images.

 'tmp/'
mode str

Wave mode to visualize ('O' or 'X').

 'O'
Source code in pynasonde/vipir/ngi/source.py 
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
def extract_ionograms(self, folder: str = "tmp/", mode: str = "O") -> None:
    """Render quick-look ionograms for each dataset to PNG files.

    Args:
        folder: Destination directory for PNG images.
        mode: Wave mode to visualize (``'O'`` or ``'X'``).
    """
    os.makedirs(folder, exist_ok=True)
    for ds in self.datasets:
        time = dt.datetime(ds.year, ds.month, ds.day, ds.hour, ds.minute, ds.second)
        i = Ionogram()
        i.add_ionogram(
            ds.Frequency, ds.Range, getattr(ds, f"{mode}_mode_power"), mode=mode
        )
        i.save(
            os.path.join(folder, f"{ds.URSI}_{time.strftime('%Y%m%d%H%M%S')}.png")
        )
        i.close()
    return

extract_FTI_RTI(rlim=[50, 800], flim=[], mode='O', noise_scale=1.0)

Extract frequency–time ionogram summary points for each dataset.

Parameters:

Name Type Description Default
rlim List[float]

Height limits (km) used when scanning for maxima.

 [50, 800]
flim List[float]

Optional frequency limits (MHz) for filtering results.

 []
mode str

Ionogram mode to process.

 'O'
noise_scale float

Multiplier applied to the noise threshold.

 1.0

Returns:

Type Description
pd.DataFrame

DataFrame of RTI points with timestamp, frequency, power, and range.
Source code in pynasonde/vipir/ngi/source.py 
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
def extract_FTI_RTI(
    self,
    rlim: List[float] = [50, 800],
    flim: List[float] = [],
    mode: str = "O",
    noise_scale: float = 1.0,
) -> pd.DataFrame:
    """Extract frequency–time ionogram summary points for each dataset.

    Args:
        rlim: Height limits (km) used when scanning for maxima.
        flim: Optional frequency limits (MHz) for filtering results.
        mode: Ionogram mode to process.
        noise_scale: Multiplier applied to the noise threshold.

    Returns:
        DataFrame of RTI points with timestamp, frequency, power, and range.
    """
    logger.info(f"Extract FTI/RTI, based on {rlim}km")
    rti = []
    for ds in self.datasets:
        time = dt.datetime(ds.year, ds.month, ds.day, ds.hour, ds.minute, ds.second)
        logger.info(f"Time: {time}")
        f_max = np.nan
        for i, r in enumerate(ds.Range):
            if (r < rlim[0]) or (r > rlim[1]):
                continue
            powr = np.array(getattr(ds, f"{mode}_mode_power")[:, i])
            powr[powr < getattr(ds, f"{mode}_mode_noise") * noise_scale] = (
                np.nan
            )  # remove noise
            f_max = ds.Frequency[np.nanargmax(powr)]
            rti.append(
                dict(
                    time=time,
                    frequency=f_max / 1e3,  # to MHz
                    power=np.nanmax(powr),  # in dB
                    range=r,
                    noise=getattr(ds, f"{mode}_mode_noise")[
                        np.nanargmax(powr)
                    ],  # in dB
                )
            )
    rti = pd.DataFrame.from_records(rti)
    if len(flim) == 2:
        rti = rti[(rti.frequency >= flim[0]) & (rti.frequency <= flim[1])]
    return rti

extract_Power_RTI(folder='tmp/', rlim=[50, 800], flim=[3.95, 4.05], mode='O', fname=None, xlabel='Time, UT', ylabel='Virtual Height, km', xlim=None, add_cbar=True, cbar_label='{}-mode Power, dB', cmap='Spectral', xtick_locator=mdates.HourLocator(interval=4), prange=[5, 70], noise_scale=1.2, date_format='$%H^{%M}$', del_ticks=False, xdate_lims=None)

Generate power RTI plots and return the underlying dataframe.

Parameters:

Name Type Description Default
folder str

Directory where the PNG plot will be written.

 'tmp/'
rlim List[float]

Range limits (km) kept in the dataset.

 [50, 800]
flim List[float]

Frequency limits (MHz) used to narrow the data.

 [3.95, 4.05]
mode str

Ionogram mode ('O' or 'X').

 'O'
fname str

Optional filename for the saved figure.

 None
xlabel str

X-axis label text.

 'Time, UT'
ylabel str

Y-axis label text.

 'Virtual Height, km'
xlim List[dt.datetime]

Datetime limits applied to the plot.

 None
add_cbar bool

Whether to include a colorbar.

 True
cbar_label str

Format string for the colorbar label.

 '{}-mode Power, dB'
cmap str

Matplotlib colormap name.

 'Spectral'
xtick_locator mdates.HourLocator

Locator controlling x-axis tick frequency.

 mdates.HourLocator(interval=4)
prange List[float]

Min/max dB levels displayed.

 [5, 70]
noise_scale float

Multiplier applied to the noise floor when masking.

 1.2
date_format str

Major tick label format.

 '$%H^{%M}$'
del_ticks bool

Remove axis ticks before plotting (useful for grids).

 False
xdate_lims List[dt.datetime]

Override for x-axis limits.

 None

Returns:

Type Description
pd.DataFrame

DataFrame containing the filtered RTI samples.
Source code in pynasonde/vipir/ngi/source.py 
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
def extract_Power_RTI(
    self,
    folder: str = "tmp/",
    rlim: List[float] = [50, 800],
    flim: List[float] = [3.95, 4.05],
    mode: str = "O",
    fname: str = None,
    xlabel: str = "Time, UT",
    ylabel: str = "Virtual Height, km",
    xlim: List[dt.datetime] = None,
    add_cbar: bool = True,
    cbar_label: str = "{}-mode Power, dB",
    cmap: str = "Spectral",
    xtick_locator: mdates.HourLocator = mdates.HourLocator(interval=4),
    prange: List[float] = [5, 70],
    noise_scale: float = 1.2,
    date_format: str = r"$%H^{%M}$",
    del_ticks: bool = False,
    xdate_lims: List[dt.datetime] = None,
) -> pd.DataFrame:
    """Generate power RTI plots and return the underlying dataframe.

    Args:
        folder: Directory where the PNG plot will be written.
        rlim: Range limits (km) kept in the dataset.
        flim: Frequency limits (MHz) used to narrow the data.
        mode: Ionogram mode (``'O'`` or ``'X'``).
        fname: Optional filename for the saved figure.
        xlabel: X-axis label text.
        ylabel: Y-axis label text.
        xlim: Datetime limits applied to the plot.
        add_cbar: Whether to include a colorbar.
        cbar_label: Format string for the colorbar label.
        cmap: Matplotlib colormap name.
        xtick_locator: Locator controlling x-axis tick frequency.
        prange: Min/max dB levels displayed.
        noise_scale: Multiplier applied to the noise floor when masking.
        date_format: Major tick label format.
        del_ticks: Remove axis ticks before plotting (useful for grids).
        xdate_lims: Override for x-axis limits.

    Returns:
        DataFrame containing the filtered RTI samples.
    """
    logger.info(f"Extract Power/RTI, based on {flim}MHz {rlim}km")
    rti = pd.DataFrame()
    for ds in self.datasets:
        time = dt.datetime(ds.year, ds.month, ds.day, ds.hour, ds.minute, ds.second)
        logger.info(f"Time: {time}")
        frequency, range = np.meshgrid(ds.Frequency, ds.Range, indexing="ij")
        noise, _ = np.meshgrid(
            getattr(ds, f"{mode}_mode_noise"), ds.Range, indexing="ij"
        )
        o = pd.DataFrame()
        (
            o["frequency"],
            o["range"],
            o[f"{mode}_mode_power"],
            o[f"{mode}_mode_noise"],
        ) = (
            frequency.ravel() / 1e3,  # to MHz
            range.ravel(),  # in km
            getattr(ds, f"{mode}_mode_power").ravel(),  # in dB
            noise.ravel(),  # in dB
        )
        o["time"] = time
        if (len(rlim) == 2) and (len(flim) == 2):
            o = o[
                (o.range >= rlim[0])
                & (o.range <= rlim[1])
                & (o.frequency >= flim[0])
                & (o.frequency <= flim[1])
            ]
        rti = pd.concat([rti, o])

    if fname is None:
        fname = f"{ds.URSI}_{rti.time.min().strftime('%Y%m%d.%H%M-')}{rti.time.max().strftime('%H%M')}_{mode}-mode.png"
    if xdate_lims is None:
        fig_title = f"""{ds.URSI}/{rti.time.min().strftime('%H%M-')}{rti.time.max().strftime('%H%M')} UT, {rti.time.max().strftime('%d %b %Y')}"""
    else:
        fig_title = f"""{ds.URSI}/{xdate_lims[0].strftime('%H%M-')}{xdate_lims[1].strftime('%H%M')} UT, {xdate_lims[1].strftime('%d %b %Y')}"""
    fig_title += (
        r"/ $f_0\sim$[" + "%.2f" % flim[0] + "-" + "%.2f" % flim[1] + "] MHz"
    )
    i = Ionogram(fig_title=fig_title, nrows=1, ncols=1)
    i.add_interval_plots(
        rti,
        mode=mode,
        xlabel=xlabel,
        ylabel=ylabel,
        xlim=xlim,
        add_cbar=add_cbar,
        cbar_label=cbar_label.format(mode),
        cmap=cmap,
        prange=prange,
        noise_scale=noise_scale,
        del_ticks=del_ticks,
        date_format=date_format,
        ylim=rlim,
        xtick_locator=xtick_locator,
        xdate_lims=xdate_lims,
    )
    i.save(os.path.join(folder, fname))
    i.close()
    return rti

save_scaled_dataset(params=['hmF2', 'foF2'])

Persist selected scaled parameters to a CSV file.

Parameters:

Name Type Description Default
params

Sequence of attribute names to extract per dataset.

 ['hmF2', 'foF2']
Source code in pynasonde/vipir/ngi/source.py 
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
def save_scaled_dataset(self, params=["hmF2", "foF2"]):
    """Persist selected scaled parameters to a CSV file.

    Args:
        params: Sequence of attribute names to extract per dataset.
    """
    records = []
    for ds in self.datasets:
        rec = dict()
        for p in params:
            rec[p] = getattr(ds, p)
        records.append(rec)
    records = pd.DataFrame.from_records(records)
    fname = os.path.join(self.source_folder, "scaled.csv")
    records.to_csv(fname, float_format="%g", header=True, index=False)
    return

save_scaled_parameters(attr=dict(), mode='O')

Write trace parameters into per-epoch NetCDF files.

Parameters:

Name Type Description Default
attr dict

Extra dataset-level attributes to store in each NetCDF file.

 dict()
mode str

Mode suffix applied to the output filenames.

 'O'
Source code in pynasonde/vipir/ngi/source.py 
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
def save_scaled_parameters(self, attr: dict = dict(), mode: str = "O"):
    """Write trace parameters into per-epoch NetCDF files.

    Args:
        attr: Extra dataset-level attributes to store in each NetCDF file.
        mode: Mode suffix applied to the output filenames.
    """
    folder = os.path.join(self.source_folder, "scaled")
    os.makedirs(folder, exist_ok=True)
    for ds in self.datasets:
        fname = os.path.join(
            folder, f"{mode}_mode_sp_{ds.time.strftime('%Y%m%d%H%M%S')}.nc"
        )
        logger.info(f"Saved to {fname}")
        max_n_trace = ds.get_n_traces()
        fs, hs = (
            np.zeros(max_n_trace) * np.nan,
            np.zeros(max_n_trace) * np.nan,
        )
        if ds.get_n_traces():
            fs, hs = (
                ds.trace.trace_params.fs,
                ds.trace.trace_params.hs,
            )
        data = {
            "fs": (("max_n_trace"), fs),
            "hs": (("max_n_trace"), hs),
            "sza": (("time",), [ds.sza]),
        }
        coords = dict(time=[ds.time], max_n_trace=np.arange(max_n_trace))
        d = xr.Dataset(data, coords=coords, attrs=attr)
        d.fs.attrs["units"], d.hs.attrs["units"] = "MHz", "km"
        d.fs.attrs["description"], d.hs.attrs["description"] = (
            "Plasma Frequencies",
            "Virtual Height/Reflection Height",
        )
        d.to_netcdf(fname)
        del d
    return