- ``smoothframes`` (default=True): Flag used when reading records with signals having multiple samples per frame. Specifies whether to smooth the samples in signals with more than one sample per frame and return an mxn uniform numpy array as the d_signals or p_signals field (True), or to return a list of 1d numpy arrays containing every expanded sample as the e_d_signals or e_p_signals field (False).

"# Demo Scripts for the wfdb-python package\n",

{

"cell_type": "code",

"execution_count": null,

"metadata": {

"collapsed": true

},

"outputs": [],

"source": [

"# See the help documentation for the read functions\n",

"\n",

"#help(wfdb.rdsamp)\n",

"#help(wfdb.srdsamp)\n",

"#help(wfdb.rdann)"

]

},

"## Reading Records and Annotations"

"### Multiple sample/frame examples\n",

"\n",

"Although there can only be one base sampling frequency per record, a single wfdb record can store multiple channels with different sampling frequencies, as long as their sampling frequencies can all be expressed by an integer multiple of a base value. This is done by using the `sampsperframe` attribute in each channel, which indicates the number of samples of each channel present in each frame.\n",

"\n",

"ie: To capture three signals with `fs = 120, 240, and 360 Hz` in a single record, they can be combined into a record with `fs = 120` and `sampsperframe = [1, 2, 3]`.\n",

"\n",

"#### Reading Options\n",

"\n",

"This package allows signals in records with multiple samples/frame to be read in two ways:\n",

"1. smoothed - An uniform mxn numpy is returned as the d_signals or p_signals field. Channels with multiple samples/frame have their values averaged within each frame. This is like the behaviour of the `rdsamp` function of the original WFDB c package. Note that `wfdb.plotrec` only works if the record object has the `p_signals` field.\n",

"2. expanded - A list of 1d numpy arrays is returned as the e_d_signals or e_p_signals field. All samples for each channel are returned in its respective numpy array. The arrays may have different lengths depending on their `sampsperframe` values. \n",

"\n",

"Set the `smoothframes` *(default=True)* option in `rdsamp` to return the desired signal type."

]

},

{

"cell_type": "code",

"execution_count": null,

"metadata": {

"collapsed": false

},

"outputs": [],

"source": [

"# Demo 8 - Read a wfdb record in which one channel has multiple samples/frame. Return a smoothed uniform array.\n",

"record = wfdb.rdsamp('sampledata/test01_00s_frame')\n",

"wfdb.plotrec(record)"

"# Demo 9 - Read a wfdb record in which one channel has multiple samples/frame. Return a list of all the expanded samples.\n",

"record = wfdb.rdsamp('sampledata/test01_00s_frame', smoothframes = False)\n",

"\n",

"display(record.e_p_signals)\n",

"# Show that different channels have different lengths. Channel 1 has 2 samples/frame, hence has 2x as many samples.\n",

"print([len(s) for s in record.e_p_signals])\n",

"\n",

"# wfdb.plotrec doesn't work because the Record object is missing its p_signals field."

]

},

{

"cell_type": "markdown",

"metadata": {},

"source": [

"## Writing Records and Annotations"

]

},

{

"cell_type": "code",

"execution_count": null,

"metadata": {

"collapsed": false

},

"outputs": [],

"source": [

"# Demo 10 - Read a WFDB record's digital samples and create a copy via the wrsamp() instance method \n",

"# Demo 11 - Write a WFDB record without using a Record object via the gateway wrsamp function.\n",

"# Demo 12 - Write a WFDB record with multiple samples/frame in a channel\n",

"\n",

"# Read a record as a Record object.\n",

"record = wfdb.rdsamp('sampledata/test01_00s_frame', physical = False, smoothframes=False)\n",

"record.recordname = 'test01_00s_framex'\n",

"\n",

"# Call the instance method of the object with expanded=True to write the record using the e_d_signals field\n",

"record.wrsamp(expanded=True)\n",

"\n",

"# The new file can be read\n",

"recordx = wfdb.rdsamp('test01_00s_framex')"

]

},

{

"cell_type": "code",

"execution_count": null,

"metadata": {

"collapsed": false

},

"outputs": [],

"source": [

"# Demo 13 - Read a WFDB annotation file and create a copy via the wrann() instance method\n",

"# Demo 14 - Write a WFDB annotation file without using an Annotator object via the gateway wrann function.\n",

"# Demo 15 - View what the 'anntype' symbols mean in the standard WFDB library\n",

"## Downloading Content from Physiobank\n",

"# Demo 16 - List the Physiobank Databases\n",

"# Demo 17 - Download all the WFDB records and annotations from a small Physiobank Database\n",

"# Demo 18 - Download specified files from a Physiobank database\n",

test01_00s_frame 3 500 4000

test01_00s.dat 16 100/mV 16 0 10 114 0 ECG_1

test01_00s.dat 16x2 100/mV 16 0 -8 822 0 ECG_2

test01_00s.dat 16 100/mV 16 0 -57 65135 0 ECG_3

# <age>: 25 <sex>: M <diagnoses>: (none) <medications>: (none)

# This is a copy of test01_00s with 2 samples/frame for channel 1, cutting out channel 3.

test01_00s_skew 4 500 4000

test01_00s.dat 16 100/mV 16 0 10 114 0 ECG_1

test01_00s.dat 16:3 100/mV 16 0 -8 941 0 ECG_2

test01_00s.dat 16 100/mV 16 0 -57 -119 0 ECG_3

test01_00s.dat 16 100/mV 16 0 -66 -401 0 ECG_4

# <age>: 25 <sex>: M <diagnoses>: (none) <medications>: (none)

# This is a copy of test01_00s with a skew stated in channel 1.

test01_00s_skewframe 3 500 4000

test01_00s.dat 16:3 100/mV 16 0 10 114 0 ECG_1

test01_00s.dat 16x2 100/mV 16 0 -8 822 0 ECG_2

test01_00s.dat 16 100/mV 16 0 -57 65135 0 ECG_3

# <age>: 25 <sex>: M <diagnoses>: (none) <medications>: (none)

# This is a copy of test01_00s with 2 samples/frame for channel 1, cutting out channel 3. A skew is also stated in signal 0.

# Checksum and initvalue to be updated if present

# unless the whole signal length was input

if self.siglen != int(len(self.e_d_signals[0])/self.sampsperframe[0]):

self.checksum = self.calc_checksum(expanded)

self.initvalue = [s[0] for s in self.e_d_signals]