Hot-keys on this page

r m x p   toggle line displays

j k   next/prev highlighted chunk

0   (zero) top of page

1   (one) first highlighted chunk

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

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

87

88

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

220

221

222

223

# -*- coding: utf-8 -*- 

""" 

Tools for creating and merging previews. 

 

:copyright: 

    The ObsPy Development Team (devs@obspy.org) 

:license: 

    GNU Lesser General Public License, Version 3 

    (http://www.gnu.org/copyleft/lesser.html) 

""" 

from copy import copy 

from obspy.core.stream import Stream 

from obspy.core.trace import Trace 

from obspy.core.utcdatetime import UTCDateTime 

import numpy as np 

 

 

def createPreview(trace, delta=60): 

    """ 

    Creates a preview trace. 

 

    A preview trace consists of maximum minus minimum of all samples within 

    ``delta`` seconds. The parameter ``delta`` must be a multiple of the 

    sampling rate of the ``trace`` object. 

 

    :type delta: integer, optional 

    :param delta: Difference between two preview points. Defaults to ``60``. 

    :rtype: :class:`~obspy.core.Trace` 

    :return: New Trace object. 

 

    This method will modify the original Trace object. Create a copy of the 

    Trace object if you want to continue using the original data. 

    """ 

    if not isinstance(delta, int) or delta < 1: 

        msg = 'The delta values need to be an Integer and at least 1.' 

        raise TypeError(msg) 

    data = trace.data 

    start_time = trace.stats.starttime.timestamp 

    # number of samples for a single slice of delta seconds 

    samples_per_slice = delta * int(trace.stats.sampling_rate) 

    if samples_per_slice < 1: 

        raise ValueError('samples_per_slice is less than 0 - skipping') 

    # minimum and maximum of samples before a static time marker 

    start = (delta - start_time % delta) * int(trace.stats.sampling_rate) 

    start_time = start_time - start_time % delta 

    if start > (delta / 2) and data[0:start].size: 

        first_diff = [data[0:start].max() - data[0:start].min()] 

    else: 

        # skip starting samples 

        first_diff = [] 

        start_time += delta 

    # number of complete slices of data 

    number_of_slices = int((len(data) - start) / samples_per_slice) 

    # minimum and maximum of remaining samples 

    end = samples_per_slice * number_of_slices + start 

    if end > (delta / 2) and data[end:].size: 

        last_diff = [data[end:].max() - data[end:].min()] 

    else: 

        # skip tailing samples 

        last_diff = [] 

    # Fill NaN value with -1. 

    if np.isnan(last_diff): 

        last_diff = -1 

    # reshape matrix 

    data = trace.data[start:end].reshape([number_of_slices, samples_per_slice]) 

    # get minimum and maximum for each row 

    diff = data.ptp(axis=1) 

    # fill masked values with -1 -> means missing data 

    if isinstance(diff, np.ma.masked_array): 

        diff = np.ma.filled(diff, -1) 

    data = np.concatenate([first_diff, diff, last_diff]) 

    data = np.require(data, dtype="float32") 

    tr = Trace(data=data, header=trace.stats) 

    tr.stats.delta = delta 

    tr.stats.npts = len(data) 

    tr.stats.starttime = UTCDateTime(start_time) 

    tr.stats.preview = True 

    return tr 

 

 

def mergePreviews(stream): 

    """ 

    Merges all preview traces in one Stream object. Does not change the 

    original stream because the data needs to be copied anyway. 

 

    :type stream: :class:`~obspy.core.Stream` 

    :param stream: Stream object to be merged 

    :rtype: :class:`~obspy.core.Stream` 

    :return: Merged Stream object. 

    """ 

    copied_traces = copy(stream.traces) 

    stream.sort() 

    # Group traces by id. 

    traces = {} 

    dtypes = [] 

    for trace in stream: 

        # Throw away empty traces. 

        if trace.stats.npts == 0: 

            continue 

        if not hasattr(trace.stats, 'preview') or not trace.stats.preview: 

            msg = 'Trace\n%s\n is no preview file.' % str(trace) 

            raise Exception(msg) 

        traces.setdefault(trace.id, []) 

        traces[trace.id].append(trace) 

        dtypes.append(trace.data.dtype) 

    if len(traces) == 0: 

        return Stream() 

    # Initialize new Stream object. 

    new_stream = Stream() 

    for value in traces.values(): 

        if len(value) == 1: 

            new_stream.append(value[0]) 

            continue 

        # All traces need to have the same delta value and also be on the same 

        # grid spacing. It is enough to only check the sampling rate because 

        # the algorithm that creates the preview assures that the grid spacing 

        # is correct. 

        sampling_rates = set([tr.stats.sampling_rate for tr in value]) 

        if len(sampling_rates) != 1: 

            msg = 'More than one sampling rate for traces with id %s.' % \ 

                  value[0].id 

            raise Exception(msg) 

        delta = value[0].stats.delta 

        # Check dtype. 

        dtypes = set([str(tr.data.dtype) for tr in value]) 

        if len(dtypes) > 1: 

            msg = 'Different dtypes for traces with id %s' % value[0].id 

            raise Exception(msg) 

        dtype = dtypes.pop() 

        # Get the minimum start and maximum endtime for all traces. 

        min_starttime = min([tr.stats.starttime for tr in value]) 

        max_endtime = max([tr.stats.endtime for tr in value]) 

        samples = (max_endtime - min_starttime) / delta + 1 

        data = np.empty(samples, dtype=dtype) 

        # Fill with negative one values which corresponds to a gap. 

        data[:] = -1 

        # Create trace and give starttime. 

        new_trace = Trace(data=data, header=value[0].stats) 

        # Loop over all traces in value and add to data. 

        for trace in value: 

            start_index = int((trace.stats.starttime - min_starttime) / delta) 

            end_index = start_index + len(trace.data) 

            # Element-by-element comparison. 

            data[start_index:end_index] = \ 

                np.maximum(data[start_index:end_index], trace.data) 

        # set npts again, because data is changed in place 

        new_trace.stats.npts = len(data) 

        new_stream.append(new_trace) 

    stream.traces = copied_traces 

    return new_stream 

 

 

def resamplePreview(trace, samples, method='accurate'): 

    """ 

    Resamples a preview Trace to the chosen number of samples. 

 

    :type trace: :class:`~obspy.core.Trace` 

    :param trace: Trace object to be resampled. 

    :type samples: int 

    :param samples: Desired number of samples. 

    :type method: str, optional 

    :param method: Resample method. Available are ``'fast'`` and 

        ``'accurate'``. Defaults to ``'accurate'``. 

 

    .. rubric:: Notes 

 

    This method will destroy the data in the original Trace object. 

    Deepcopy the Trace if you want to continue using the original data. 

 

    The fast method works by reshaping the data array to a 

    sample x int(npts/samples) matrix (npts are the number of samples in 

    the original trace) and taking the maximum of each row. Therefore 

    the last npts - int(npts/samples)*samples will be omitted. The worst 

    case scenario is resampling a 1999 samples array to 1000 samples. 999 

    samples, almost half the data will be omitted. 

 

    The accurate method has no such problems because it will move a window 

    over the whole array and take the maximum for each window. It loops 

    over each window and is up to 10 times slower than the fast method. 

    This of course is highly depended on the number of wished samples and 

    the original trace and usually the accurate method is still fast 

    enough. 

    """ 

    # Only works for preview traces. 

    if not hasattr(trace.stats, 'preview') or not trace.stats.preview: 

        msg = 'Trace\n%s\n is no preview file.' % str(trace) 

        raise Exception(msg) 

    # Save same attributes for later use. 

    endtime = trace.stats.endtime 

    dtype = trace.data.dtype 

    npts = trace.stats.npts 

    # XXX: Interpolate? 

    if trace.stats.npts < samples: 

        msg = 'Can only downsample so far. Interpolation not yet implemented.' 

        raise NotImplementedError(msg) 

    # Return if no change is necessary. There obviously are no omitted samples. 

    elif trace.stats.npts == samples: 

        return 0 

    # Fast method. 

    if method == 'fast': 

        trace.data = trace.data[:int(npts / samples) * samples] 

        trace.data = trace.data.reshape(samples, len(trace.data) // samples) 

        trace.data = trace.data.max(axis=1) 

        # Set new sampling rate. 

        trace.stats.delta = (endtime - trace.stats.starttime) / \ 

            float(samples - 1) 

        # Return number of omitted samples. 

        return npts - int(npts / samples) * samples 

    # Slow but accurate method. 

    elif method == 'accurate': 

        new_data = np.empty(samples, dtype=dtype) 

        step = trace.stats.npts / float(samples) 

        for _i in xrange(samples): 

            new_data[_i] = trace.data[int(_i * step): 

                                      int((_i + 1) * step)].max() 

        trace.data = new_data 

        # Set new sampling rate. 

        trace.stats.delta = (endtime - trace.stats.starttime) / \ 

            float(samples - 1) 

        # Return number of omitted samples. Should be 0 for this method. 

        return npts - int(samples * step) 

    else: 

        raise NotImplementedError('Unknown method')