Merge pull request #146 from mhearne-usgs/depfix

Cleaned up command line behavior a bit, modified origin description p…

Author: Mike Hearne

bin/geteventhist

@@ -9,23 +9,31 @@ import textwrap
 import pandas as pd
 from openpyxl import load_workbook
 from obspy.geodetics.base import gps2dist_azimuth
+import numpy as np
 
 # local imports
 from libcomcat.search import search
 from libcomcat.dataframes import (get_history_data_frame, split_history_frame,
                                   PRODUCTS, TIMEFMT, PRODUCT_COLUMNS)
 
 
+DISPLAY_TIME_FMT = '%Y-%m-%d %H:%M:%S'
+
+
+class MyFormatter(argparse.RawTextHelpFormatter,
+                  argparse.ArgumentDefaultsHelpFormatter):
+    pass
+
+
 def get_parser():
     desc = '''Print out ComCat event history.
 
     Usage:
         geteventhist ci38457511
 
     '''
-    formatter = argparse.RawDescriptionHelpFormatter
     parser = argparse.ArgumentParser(description=desc,
-                                     formatter_class=formatter)
+                                     formatter_class=MyFormatter)
     # positional arguments
     parser.add_argument('eventid',
                         metavar='EVENTID', help='ComCat event ID.')
@@ -53,7 +61,7 @@ def get_parser():
     parser.add_argument('-o', '--outdir', help=ohelp,
                         default=None)
 
-    parser.add_argument('-w', '--web', help='Produce HTML output.',
+    parser.add_argument('-w', '--web', help='Print HTML tables to stdout.',
                         default=False, action='store_true')
 
     parser.add_argument('-f', '--format', help='Control output format',
@@ -113,10 +121,10 @@ def save_dataframe(outdir, format, event, dataframe, product=None):
         wb.save(outfile)
     else:
         if product is not None:
-            outfile = os.path.join(args.outdir,
-                                   event.id + '_' + products[0] + '.csv')
+            outfile = os.path.join(outdir,
+                                   event.id + '_' + product + '.csv')
         else:
-            outfile = os.path.join(args.outdir, event.id + '.csv')
+            outfile = os.path.join(outdir, event.id + '.csv')
         dataframe.to_csv(outfile, index=False)
         cdata = open(outfile, 'rt').read()
         with open(outfile, 'wt') as f:
@@ -131,10 +139,43 @@ def save_dataframe(outdir, format, event, dataframe, product=None):
     return outfile
 
 
+def web_print(event, dataframe):
+    etable_fmt = '''
+    <pre>
+    Event ID: %s
+    Origin Time: %s
+    Magnitude: %.1f
+    Latitude: %.4f
+    Longitude: %.4f
+    Depth: %.1f
+    </pre>
+    '''
+    etable_tpl = (event.id,
+                  event.time.strftime(TIMEFMT),
+                  event.magnitude,
+                  event.latitude,
+                  event.longitude,
+                  event.depth)
+    etable = etable_fmt % etable_tpl
+    print(textwrap.dedent(etable))
+    print(dataframe.to_html(index=False, border=0, max_rows=None, max_cols=None))
+
+
+def simplify_times(dataframe):
+    # re-format all time columns to be like: 2019-01-01 17:34:16.1
+    # first figure out all the time-like columns
+    # df['date'] = pd.to_datetime(df["date"].dt.strftime('%Y-%m'))
+    dtypes = dataframe.dtypes
+    for idx, dtype in dtypes.iteritems():
+        if np.issubdtype(dtype, np.datetime64):
+            dataframe[idx] = pd.to_datetime(
+                dataframe[idx].dt.strftime(DISPLAY_TIME_FMT))
+
+
 def main():
     pd.set_option('display.width', 1000)
     pd.set_option('display.max_rows', 1000)
-    pd.set_option('display.max_colwidth', 120)
+    pd.set_option('display.max_colwidth', -1)
     pd.set_option('display.max_columns', 1000)
     pd.set_option("display.colheader_justify", "left")
     parser = get_parser()
@@ -155,6 +196,16 @@ def main():
         print(fmt % (','.join(unsupported)))
         sys.exit(1)
 
+    # web output and directory output are mutually exclusive
+    if args.outdir and args.web:
+        msg = '''The -o and -w options are mutually exclusive, meaning
+        that you cannot choose to write files to a directory and print
+        HTML output to the screen simultaneously. Please choose one of
+        those two options and try again.
+        '''
+        print(msg)
+        sys.exit(1)
+
     if args.products:
         products = args.products
     else:
@@ -196,8 +247,14 @@ def main():
         # now re-sort by update time
         dataframe = dataframe.sort_values('Update Time')
         dataframe = dataframe[PRODUCT_COLUMNS]
+    else:
+        # since "Authoritative Event ID" and "Associated" columns are only applicable when
+        # we're including other events in our results, drop those columns
+        # if we're not doing that.
+        drop_columns = ['Authoritative Event ID', 'Associated']
+        dataframe = dataframe.drop(drop_columns, axis='columns')
 
-    if not os.path.isdir(args.outdir):
+    if args.outdir is not None and not os.path.isdir(args.outdir):
         os.makedirs(args.outdir)
 
     if args.split:
@@ -207,36 +264,23 @@ def main():
         # somehow in this process
         for product in available_products:
             pframe = split_history_frame(dataframe, product=product)
-            outfile = save_dataframe(args.outdir, args.format, event,
-                                     pframe, product=product)
-            print('%i rows saved to %s' % (len(pframe), outfile))
+            simplify_times(pframe)
+            if args.web:
+                web_print(event, pframe)
+            else:
+                outfile = save_dataframe(args.outdir, args.format, event,
+                                         pframe, product=product)
+                print('%i rows saved to %s' % (len(pframe), outfile))
+        sys.exit(0)
+
     if args.outdir:
         outfile = save_dataframe(args.outdir, args.format, event,
                                  dataframe, product=None)
 
         print('%i rows saved to %s' % (len(dataframe), outfile))
     elif args.web:
-        etable_fmt = '''
-        <pre>
-        Event ID: %s
-        Origin Time: %s
-        Magnitude: %.1f
-        Latitude: %.4f
-        Longitude: %.4f
-        Depth: %.1f
-        </pre>
-        '''
-        etable_tpl = (event.id,
-                      event.time.strftime(TIMEFMT),
-                      event.magnitude,
-                      event.latitude,
-                      event.longitude,
-                      event.depth)
-        etable = etable_fmt % etable_tpl
-        print(textwrap.dedent(etable))
-        print(dataframe.to_html(index=False, border=0))
-    else:
-        print(dataframe.to_string(index=False))
+        simplify_times(dataframe)
+        web_print(event, dataframe)
 
     sys.exit(0)
 

libcomcat/dataframes.py

@@ -16,6 +16,7 @@
 import requests
 from scipy.special import erfcinv
 from obspy.geodetics.base import gps2dist_azimuth
+from impactutils.mapping.compass import get_compass_dir_azimuth
 
 # local imports
 from libcomcat.classes import VersionOption
@@ -54,9 +55,9 @@
     'Hypocentral distance': 'distance'
 }
 
-PRODUCT_COLUMNS = ['Product', 'Authoritative Event ID', 'Code', 'Associated',
+PRODUCT_COLUMNS = ['Update Time', 'Product', 'Authoritative Event ID', 'Code', 'Associated',
                    'Product Source', 'Product Version',
-                   'Update Time', 'Elapsed (min)', 'Description']
+                   'Elapsed (min)', 'Description']
 
 SECSPERDAY = 86400
 TIMEFMT = '%Y-%m-%d %H:%M:%S'
@@ -1284,7 +1285,7 @@ def _describe_origin(event, product):
     if product.hasProperty('magnitude-type'):
         magtype = product['magnitude-type']
 
-    otime = 'unknown'
+    otime = 'NaT'
     tdiff = np.nan
     if product.hasProperty('eventtime'):
         otime_str = product['eventtime']
@@ -1295,26 +1296,47 @@ def _describe_origin(event, product):
     elapsed_sec = elapsed.days * SECSPERDAY + \
         elapsed.seconds + elapsed.microseconds / 1e6
     elapsed_min = elapsed_sec / 60
+
     olat = np.nan
     olon = np.nan
+    odepth = np.nan
     dist = np.nan
-
+    az = np.nan
     if product.hasProperty('latitude'):
         olat = float(product['latitude'])
         olon = float(product['longitude'])
         odepth = float(product['depth'])
-        dist_m, _, _ = gps2dist_azimuth(authlat, authlon, olat, olon)
+        dist_m, az, _ = gps2dist_azimuth(authlat, authlon, olat, olon)
         dist = dist_m / 1000.0
+    else:
+        if len(product.getContentsMatching('quakeml.xml')):
+            unpickler = Unpickler()
+            cbytes, url = product.getContentBytes('quakeml.xml')
+            catalog = unpickler.loads(cbytes)
+            evt = catalog.events[0]
+            if hasattr(evt, 'origin'):
+                origin = evt.origin
+                olat = origin.latitude
+                olon = origin.longitude
+                odepth = origin.depth/1000
+                dist_m, az, _ = gps2dist_azimuth(authlat, authlon, olat, olon)
+                dist = dist_m / 1000.0
 
     loc_method = 'unknown'
     if product.hasProperty('cube-location-method'):
         loc_method = product['cube-location-method']
 
-    fmt = ('Magnitude# %.1f|Depth# %.1f|Time# %s |Time Offset (sec)# %.1f|'
+    # convert azimuth to cardinal direction
+    if not np.isnan(az):
+        azstr = get_compass_dir_azimuth(az, resolution='meteorological')
+    else:
+        azstr = ''
+
+    fmt = ('Magnitude# %.1f|Time# %s |Time Offset (sec)# %.1f|'
            'Location# (%.3f,%.3f)|Distance from Auth. Origin (km)# %.1f|'
-           'Magnitude Type# %s|Location Method# %s')
-    desc = fmt % (omag, odepth, otime, tdiff,
-                  olat, olon, dist, magtype, loc_method)
+           'Azimuth# %s|Depth# %.1f|Magnitude Type# %s|Location Method# %s')
+    desc = fmt % (omag, otime, tdiff,
+                  olat, olon, dist, azstr, odepth, magtype, loc_method)
     productsource = product['eventsource']
     pversion = product.version
     row = {'Product': product.name,
@@ -1630,15 +1652,14 @@ def split_history_frame(dataframe, product=None):
                 newval = float(val)
             except ValueError:
                 try:
-                    newval = datetime.strptime(val, TIMEFMT3)
+                    newval = pd.Timestamp(val)
                 except ValueError:
                     newval = val
             newvalues.append(newval)
         ddict = dict(zip(columns, newvalues))
         row = pd.Series(ddict)
         df2 = df2.append(row, ignore_index=True)
 
-    # todo: something weird happening in concat step
     dataframe = dataframe.reset_index(drop=True)
     df2 = df2.reset_index(drop=True)
     dataframe = pd.concat([dataframe, df2], axis=1)