[GitHub] rhunwicks commented on a change in pull request #3492: PandasConnector
rhunwicks commented on a change in pull request #3492: PandasConnector
URL:
https://github.com/apache/incubator-superset/pull/3492#discussion_r142071323
##
File path: dev-reqs.txt
##
@@ -1,7 +1,10 @@
+beautifulsoup4==4.6.0
+lxml==3.8.0
codeclimate-test-reporter
coveralls
flake8
flask_cors
+lxml==3.8.0
Review comment:
I have added the necessary packages to `setup.py` so that `PandasDatasource`
works for csv, html, json and Excel.
@xrmx do you think it would be a good idea to allow
`PandasDatasource.FORMATS` to be overridden from settings, much as we allow
`ADDITIONAL_MODULE_DS_MAP` to add extra Connectors? Then a specific
installation could, for example, install `pytables` and then set:
```python
ADDITIONAL_PANDAS_FORMATS = [('hdf', 'HDF5')]
```
Another approach would be to attempt to dynamically set the list of
allowable formats based on whether we can import the required dependencies.
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] rhunwicks commented on a change in pull request #3492: PandasConnector
rhunwicks commented on a change in pull request #3492: PandasConnector URL: https://github.com/apache/incubator-superset/pull/3492#discussion_r141795500 ## File path: dev-reqs.txt ## @@ -1,7 +1,10 @@ +beautifulsoup4==4.6.0 +lxml==3.8.0 codeclimate-test-reporter coveralls flake8 flask_cors +lxml==3.8.0 Review comment: I agree that Arrow is the future. http://wesmckinney.com/blog/apache-arrow-pandas-internals/ is a good read. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] rhunwicks commented on a change in pull request #3492: PandasConnector
rhunwicks commented on a change in pull request #3492: PandasConnector URL: https://github.com/apache/incubator-superset/pull/3492#discussion_r141713559 ## File path: dev-reqs.txt ## @@ -1,7 +1,10 @@ +beautifulsoup4==4.6.0 +lxml==3.8.0 codeclimate-test-reporter coveralls flake8 flask_cors +lxml==3.8.0 Review comment: @xrmx - you are correct - we are specifically supporting html in `FORMATS` so I need to add those dependencies to `setup.py` or whatever requirements file we ship with PandasConnector if it lives outside core. I wonder if there are any other formats we should support by default? JSON should work by default. Excel might be useful for many people and would need `xlrd`? Given the normal Superset audience do we want to support HDF5 and/or Feather? What about SAS or Stata? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] rhunwicks commented on a change in pull request #3492: PandasConnector
rhunwicks commented on a change in pull request #3492: PandasConnector URL: https://github.com/apache/incubator-superset/pull/3492#discussion_r141663500 ## File path: dev-reqs.txt ## @@ -1,7 +1,10 @@ +beautifulsoup4==4.6.0 +lxml==3.8.0 codeclimate-test-reporter coveralls flake8 flask_cors +lxml==3.8.0 Review comment: > The original comment was because you added lxml twice Thanks - I've fixed it now. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] rhunwicks commented on a change in pull request #3492: PandasConnector
rhunwicks commented on a change in pull request #3492: PandasConnector URL: https://github.com/apache/incubator-superset/pull/3492#discussion_r141656438 ## File path: dev-reqs.txt ## @@ -1,7 +1,10 @@ +beautifulsoup4==4.6.0 +lxml==3.8.0 codeclimate-test-reporter coveralls flake8 flask_cors +lxml==3.8.0 Review comment: https://pandas.pydata.org/pandas-docs/stable/io.html#io-html-gotchas explains that `read_html` uses bs4 as a wrapper for lxml. We could use `html5lib` instead of `lxml`. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] rhunwicks commented on a change in pull request #3492: PandasConnector
rhunwicks commented on a change in pull request #3492: PandasConnector URL: https://github.com/apache/incubator-superset/pull/3492#discussion_r141652942 ## File path: dev-reqs.txt ## @@ -1,7 +1,10 @@ +beautifulsoup4==4.6.0 +lxml==3.8.0 codeclimate-test-reporter coveralls flake8 flask_cors +lxml==3.8.0 Review comment: They are required by `pandas.read_html()`, which I use in `connector_tests` to make it easy to set up test data as a Markdown table and then push it into whatever Datasource is being tested. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] rhunwicks commented on a change in pull request #3492: PandasConnector
rhunwicks commented on a change in pull request #3492: PandasConnector
URL:
https://github.com/apache/incubator-superset/pull/3492#discussion_r139944516
##
File path: contrib/connectors/pandas/models.py
##
@@ -0,0 +1,705 @@
+from collections import OrderedDict
+from datetime import datetime
+import logging
+from past.builtins import basestring
+try:
+from urllib.parse import urlparse
+except ImportError:
+from urlparse import urlparse
+
+import pandas as pd
+from pandas.api.types import (
+is_string_dtype, is_numeric_dtype, is_datetime64_any_dtype)
+
+from sqlalchemy import (
+Column, Integer, String, ForeignKey, Text
+)
+import sqlalchemy as sa
+from sqlalchemy.orm import backref, relationship
+from sqlalchemy_utils import ChoiceType, JSONType
+
+from flask import escape, Markup
+from flask_appbuilder import Model
+from flask_babel import lazy_gettext as _
+
+from superset import db, utils, sm
+from superset.connectors.base.models import (
+BaseDatasource, BaseColumn, BaseMetric)
+from superset.models.helpers import QueryResult, set_perm
+from superset.utils import QueryStatus
+
+
+class PandasDatabase(object):
+"""Non-ORM object for a Pandas Source"""
+
+def __init__(self, database_name, cache_timeout=None):
+self.database_name = database_name
+self.cache_timeout = cache_timeout
+
+def __str__(self):
+return self.database_name
+
+
+class PandasColumn(Model, BaseColumn):
+"""
+ORM object for Pandas columns.
+
+Each Pandas Datasource can have multiple columns"""
+
+__tablename__ = 'pandas_columns'
+
+id = Column(Integer, primary_key=True)
+pandas_datasource_id = Column(Integer, ForeignKey('pandas_datasources.id'))
+datasource = relationship(
+'PandasDatasource',
+backref=backref('columns', cascade='all, delete-orphan'),
+foreign_keys=[pandas_datasource_id])
+
+@property
+def is_num(self):
+return self.type and is_numeric_dtype(self.type)
+
+@property
+def is_time(self):
+return self.type and is_datetime64_any_dtype(self.type)
+
+@property
+def is_dttm(self):
+return self.is_time
+
+@property
+def is_string(self):
+return self.type and is_string_dtype(self.type)
+
+num_types = (
+'DOUBLE', 'FLOAT', 'INT', 'BIGINT',
+'LONG', 'REAL', 'NUMERIC', 'DECIMAL'
+)
+date_types = ('DATE', 'TIME', 'DATETIME')
+str_types = ('VARCHAR', 'STRING', 'CHAR')
+
+@property
+def expression(self):
+return ''
+
+@property
+def data(self):
+attrs = (
+'column_name', 'verbose_name', 'description', 'expression',
+'filterable', 'groupby')
+return {s: getattr(self, s) for s in attrs}
+
+
+class PandasMetric(Model, BaseMetric):
+"""
+ORM object for Pandas metrics.
+
+Each Pandas Datasource can have multiple metrics
+"""
+
+__tablename__ = 'pandas_metrics'
+
+id = Column(Integer, primary_key=True)
+pandas_datasource_id = Column(Integer, ForeignKey('pandas_datasources.id'))
+datasource = relationship(
+'PandasDatasource',
+backref=backref('metrics', cascade='all, delete-orphan'),
+foreign_keys=[pandas_datasource_id])
+source = Column(Text)
+expression = Column(Text)
+
+@property
+def perm(self):
+if self.datasource:
+return ('{parent_name}.[{obj.metric_name}]'
+'(id:{obj.id})').format(
+obj=self,
+parent_name=self.datasource.full_name)
+return None
+
+
+class PandasDatasource(Model, BaseDatasource):
+"""A datasource based on a Pandas DataFrame"""
+
+FORMATS = [
+('csv', 'CSV'),
+('html', 'HTML')
+]
+
+# See
http://pandas.pydata.org/pandas-docs/stable/timeseries.html#offset-aliases #
NOQA
+GRAINS = OrderedDict([
+('5 seconds', '5S'),
+('30 seconds', '30S'),
+('1 minute', 'T'),
+('5 minutes', '5T'),
+('1 hour', 'H'),
+('6 hour', '6H'),
+('day', 'D'),
+('one day', 'D'),
+('1 day', 'D'),
+('7 days', '7D'),
+('week', 'W-MON'),
+('week_starting_sunday', 'W-SUN'),
+('week_ending_saturday', 'W-SUN'),
+('month', 'M'),
+('quarter', 'Q'),
+('year', 'A'),
+])
+
+__tablename__ = 'pandas_datasources'
+type = 'pandas'
+baselink = 'pandasdatasourcemodelview' # url portion pointing to
ModelView endpoint
+column_class = PandasColumn
+metric_class = PandasMetric
+
+name = Column(String(100), nullable=False)
+source_url = Column(String(1000), nullable=False)
+format = Column(String(20), nullable=False)
+additional_parameters = Column(JSONType)
+
+user_id = Column(Integer, ForeignKey('ab_user.id'))
+owner = relationship(
+sm.user_model,
+backref='pandas_datasources',
+
[GitHub] rhunwicks commented on a change in pull request #3492: PandasConnector
rhunwicks commented on a change in pull request #3492: PandasConnector
URL:
https://github.com/apache/incubator-superset/pull/3492#discussion_r139937218
##
File path: contrib/connectors/pandas/views.py
##
@@ -0,0 +1,270 @@
+"""Views used by the SqlAlchemy connector"""
+import logging
+
+from past.builtins import basestring
+
+from flask import Markup, flash, redirect
+from flask_appbuilder import CompactCRUDMixin, expose
+from flask_appbuilder.models.sqla.interface import SQLAInterface
+import sqlalchemy as sa
+
+from flask_babel import lazy_gettext as _
+from flask_babel import gettext as __
+
+from superset import appbuilder, db, utils, security, sm
+from superset.utils import has_access
+from superset.connectors.base.views import DatasourceModelView
+from superset.views.base import (
+SupersetModelView, ListWidgetWithCheckboxes, DeleteMixin, DatasourceFilter,
+get_datasource_exist_error_mgs,
+)
+
+from . import models
+
+
+class PandasColumnInlineView(CompactCRUDMixin, SupersetModelView): # noqa
+datamodel = SQLAInterface(models.PandasColumn)
+
+list_title = _('List Columns')
+show_title = _('Show Column')
+add_title = _('Add Column')
+edit_title = _('Edit Column')
+
+can_delete = False
+list_widget = ListWidgetWithCheckboxes
+edit_columns = [
+'column_name', 'verbose_name', 'description',
+'type', 'groupby', 'filterable',
+'datasource', 'count_distinct', 'sum', 'min', 'max']
+add_columns = edit_columns
+list_columns = [
+'column_name', 'verbose_name', 'type', 'groupby', 'filterable',
+'count_distinct', 'sum', 'min', 'max']
+page_size = 500
+description_columns = {
+'is_dttm': _(
+"Whether to make this column available as a "
+"[Time Granularity] option, column has to be DATETIME or "
+"DATETIME-like"),
+'filterable': _(
+"Whether this column is exposed in the `Filters` section "
+"of the explore view."),
+'type': _(
+"The data type that was inferred by Pandas. "
+"It may be necessary to input a type manually for "
+"expression-defined columns in some cases. In most case "
+"users should not need to alter this."),
+}
+label_columns = {
+'column_name': _("Column"),
+'verbose_name': _("Verbose Name"),
+'description': _("Description"),
+'groupby': _("Groupable"),
+'filterable': _("Filterable"),
+'datasource': _("Datasource"),
+'count_distinct': _("Count Distinct"),
+'sum': _("Sum"),
+'min': _("Min"),
+'max': _("Max"),
+'type': _('Type'),
+}
+
+
+appbuilder.add_view_no_menu(PandasColumnInlineView)
+
+
+class PandasMetricInlineView(CompactCRUDMixin, SupersetModelView): # noqa
+datamodel = SQLAInterface(models.PandasMetric)
+
+list_title = _('List Metrics')
+show_title = _('Show Metric')
+add_title = _('Add Metric')
+edit_title = _('Edit Metric')
+
+list_columns = ['metric_name', 'verbose_name', 'metric_type']
+edit_columns = [
+'metric_name', 'description', 'verbose_name', 'metric_type',
+'source', 'expression', 'datasource', 'd3format', 'is_restricted']
+description_columns = {
+'source': utils.markdown(
+"a comma-separated list of column(s) used to calculate "
+" the metric. Example: `claim_amount`", True),
+'expression': utils.markdown(
+"a valid Pandas expression as supported by the underlying "
+"backend. Example: `count()`", True),
+'is_restricted': _("Whether the access to this metric is restricted "
+ "to certain roles. Only roles with the permission "
+ "'metric access on XXX (the name of this metric)' "
+ "are allowed to access this metric"),
+'d3format': utils.markdown(
+"d3 formatting string as defined [here]"
+"(https://github.com/d3/d3-format/blob/master/README.md#format). "
+"For instance, this default formatting applies in the Table "
+"visualization and allow for different metric to use different "
+"formats", True
+),
+}
+add_columns = edit_columns
+page_size = 500
+label_columns = {
+'metric_name': _("Metric"),
+'description': _("Description"),
+'verbose_name': _("Verbose Name"),
+'metric_type': _("Type"),
+'source': _("Pandas Source Columns"),
+'expression': _("Pandas Expression"),
+'datasource': _("Datasource"),
+'d3format': _("D3 Format"),
+'is_restricted': _('Is Restricted')
+}
+
+def post_add(self, metric):
+if metric.is_restricted:
+security.merge_perm(sm, 'metric_access', metric.get_perm())
+
+def post_update(self, metric):
+if
[GitHub] rhunwicks commented on a change in pull request #3492: PandasConnector
rhunwicks commented on a change in pull request #3492: PandasConnector
URL:
https://github.com/apache/incubator-superset/pull/3492#discussion_r139931452
##
File path: contrib/connectors/pandas/models.py
##
@@ -0,0 +1,724 @@
+from collections import OrderedDict
+from datetime import datetime
+import logging
+from past.builtins import basestring
+try:
+from urllib.parse import urlparse
+except ImportError:
+from urlparse import urlparse
+
+import pandas as pd
+from pandas.api.types import (
+is_string_dtype, is_numeric_dtype, is_datetime64_any_dtype)
+
+from sqlalchemy import (
+Column, Integer, String, ForeignKey, Text
+)
+import sqlalchemy as sa
+from sqlalchemy.orm import backref, relationship
+from sqlalchemy_utils import ChoiceType, JSONType
+
+from flask import escape, Markup
+from flask_appbuilder import Model
+from flask_babel import lazy_gettext as _
+
+from superset import db, utils, sm
+from superset.connectors.base.models import (
+BaseDatasource, BaseColumn, BaseMetric)
+from superset.models.helpers import QueryResult, set_perm
+from superset.utils import QueryStatus
+
+
+class PandasDatabase(object):
+"""Non-ORM object for a Pandas Source"""
+database_name = ''
+
+cache_timeout = None
+
+def __init__(self, database_name, cache_timeout):
+self.database_name = database_name
+self.cache_timeout = cache_timeout
+
+def __str__(self):
+return self.database_name
+
+
+class PandasColumn(Model, BaseColumn):
+"""
+ORM object for Pandas columns.
+
+Each Pandas Datasource can have multiple columns"""
+
+__tablename__ = 'pandascolumns'
+
+id = Column(Integer, primary_key=True)
+pandasdatasource_id = Column(Integer, ForeignKey('pandasdatasources.id'))
+datasource = relationship(
+'PandasDatasource',
+backref=backref('columns', cascade='all, delete-orphan'),
+foreign_keys=[pandasdatasource_id])
+
+@property
+def is_num(self):
+return self.type and is_numeric_dtype(self.type)
+
+@property
+def is_time(self):
+return self.type and is_datetime64_any_dtype(self.type)
+
+@property
+def is_dttm(self):
+return self.is_time
+
+@property
+def is_string(self):
+return self.type and is_string_dtype(self.type)
+
+num_types = (
+'DOUBLE', 'FLOAT', 'INT', 'BIGINT',
+'LONG', 'REAL', 'NUMERIC', 'DECIMAL'
+)
+date_types = ('DATE', 'TIME', 'DATETIME')
+str_types = ('VARCHAR', 'STRING', 'CHAR')
+
+@property
+def expression(self):
+return ''
+
+@property
+def data(self):
+attrs = (
+'column_name', 'verbose_name', 'description', 'expression',
+'filterable', 'groupby')
+return {s: getattr(self, s) for s in attrs}
+
+
+class PandasMetric(Model, BaseMetric):
+"""
+ORM object for Pandas metrics.
+
+Each Pandas Datasource can have multiple metrics
+"""
+
+__tablename__ = 'pandasmetrics'
+
+id = Column(Integer, primary_key=True)
+pandasdatasource_id = Column(Integer, ForeignKey('pandasdatasources.id'))
+datasource = relationship(
+'PandasDatasource',
+backref=backref('metrics', cascade='all, delete-orphan'),
+foreign_keys=[pandasdatasource_id])
+source = Column(Text)
+expression = Column(Text)
+
+@property
+def perm(self):
+if self.datasource:
+return ('{parent_name}.[{obj.metric_name}]'
+'(id:{obj.id})').format(
+obj=self,
+parent_name=self.datasource.full_name)
+return None
+
+
+class PandasDatasource(Model, BaseDatasource):
+"""A datasource based on a Pandas DataFrame"""
+
+FORMATS = [
+('csv', 'CSV'),
+('html', 'HTML')
+]
+
+# See
http://pandas.pydata.org/pandas-docs/stable/timeseries.html#offset-aliases #
NOQA
+GRAINS = OrderedDict([
+('5 seconds', '5S'),
+('30 seconds', '30S'),
+('1 minute', 'T'),
+('5 minutes', '5T'),
+('1 hour', 'H'),
+('6 hour', '6H'),
+('day', 'D'),
+('one day', 'D'),
+('1 day', 'D'),
+('7 days', '7D'),
+('week', 'W-MON'),
+('week_starting_sunday', 'W-SUN'),
+('week_ending_saturday', 'W-SUN'),
+('month', 'M'),
+('quarter', 'Q'),
+('year', 'A'),
+])
+
+__tablename__ = 'pandasdatasources'
+type = 'pandas'
+baselink = 'pandasdatasourcemodelview' # url portion pointing to
ModelView endpoint
+column_class = PandasColumn
+metric_class = PandasMetric
+
+name = Column(String(100), nullable=False)
+source_url = Column(String(1000), nullable=False)
+format = Column(String(20), nullable=False)
+additional_parameters = Column(JSONType)
+
+user_id = Column(Integer, ForeignKey('ab_user.id'))
+owner = relationship(
+sm.user_model,
+
[GitHub] rhunwicks commented on a change in pull request #3492: PandasConnector
rhunwicks commented on a change in pull request #3492: PandasConnector
URL:
https://github.com/apache/incubator-superset/pull/3492#discussion_r139931531
##
File path: contrib/connectors/pandas/models.py
##
@@ -0,0 +1,724 @@
+from collections import OrderedDict
+from datetime import datetime
+import logging
+from past.builtins import basestring
+try:
+from urllib.parse import urlparse
+except ImportError:
+from urlparse import urlparse
+
+import pandas as pd
+from pandas.api.types import (
+is_string_dtype, is_numeric_dtype, is_datetime64_any_dtype)
+
+from sqlalchemy import (
+Column, Integer, String, ForeignKey, Text
+)
+import sqlalchemy as sa
+from sqlalchemy.orm import backref, relationship
+from sqlalchemy_utils import ChoiceType, JSONType
+
+from flask import escape, Markup
+from flask_appbuilder import Model
+from flask_babel import lazy_gettext as _
+
+from superset import db, utils, sm
+from superset.connectors.base.models import (
+BaseDatasource, BaseColumn, BaseMetric)
+from superset.models.helpers import QueryResult, set_perm
+from superset.utils import QueryStatus
+
+
+class PandasDatabase(object):
+"""Non-ORM object for a Pandas Source"""
+database_name = ''
+
+cache_timeout = None
+
+def __init__(self, database_name, cache_timeout):
+self.database_name = database_name
+self.cache_timeout = cache_timeout
+
+def __str__(self):
+return self.database_name
+
+
+class PandasColumn(Model, BaseColumn):
+"""
+ORM object for Pandas columns.
+
+Each Pandas Datasource can have multiple columns"""
+
+__tablename__ = 'pandascolumns'
+
+id = Column(Integer, primary_key=True)
+pandasdatasource_id = Column(Integer, ForeignKey('pandasdatasources.id'))
+datasource = relationship(
+'PandasDatasource',
+backref=backref('columns', cascade='all, delete-orphan'),
+foreign_keys=[pandasdatasource_id])
+
+@property
+def is_num(self):
+return self.type and is_numeric_dtype(self.type)
+
+@property
+def is_time(self):
+return self.type and is_datetime64_any_dtype(self.type)
+
+@property
+def is_dttm(self):
+return self.is_time
+
+@property
+def is_string(self):
+return self.type and is_string_dtype(self.type)
+
+num_types = (
+'DOUBLE', 'FLOAT', 'INT', 'BIGINT',
+'LONG', 'REAL', 'NUMERIC', 'DECIMAL'
+)
+date_types = ('DATE', 'TIME', 'DATETIME')
+str_types = ('VARCHAR', 'STRING', 'CHAR')
+
+@property
+def expression(self):
+return ''
+
+@property
+def data(self):
+attrs = (
+'column_name', 'verbose_name', 'description', 'expression',
+'filterable', 'groupby')
+return {s: getattr(self, s) for s in attrs}
+
+
+class PandasMetric(Model, BaseMetric):
+"""
+ORM object for Pandas metrics.
+
+Each Pandas Datasource can have multiple metrics
+"""
+
+__tablename__ = 'pandasmetrics'
+
+id = Column(Integer, primary_key=True)
+pandasdatasource_id = Column(Integer, ForeignKey('pandasdatasources.id'))
+datasource = relationship(
+'PandasDatasource',
+backref=backref('metrics', cascade='all, delete-orphan'),
+foreign_keys=[pandasdatasource_id])
+source = Column(Text)
+expression = Column(Text)
+
+@property
+def perm(self):
+if self.datasource:
+return ('{parent_name}.[{obj.metric_name}]'
+'(id:{obj.id})').format(
+obj=self,
+parent_name=self.datasource.full_name)
+return None
+
+
+class PandasDatasource(Model, BaseDatasource):
+"""A datasource based on a Pandas DataFrame"""
+
+FORMATS = [
+('csv', 'CSV'),
+('html', 'HTML')
+]
+
+# See
http://pandas.pydata.org/pandas-docs/stable/timeseries.html#offset-aliases #
NOQA
+GRAINS = OrderedDict([
+('5 seconds', '5S'),
+('30 seconds', '30S'),
+('1 minute', 'T'),
+('5 minutes', '5T'),
+('1 hour', 'H'),
+('6 hour', '6H'),
+('day', 'D'),
+('one day', 'D'),
+('1 day', 'D'),
+('7 days', '7D'),
+('week', 'W-MON'),
+('week_starting_sunday', 'W-SUN'),
+('week_ending_saturday', 'W-SUN'),
+('month', 'M'),
+('quarter', 'Q'),
+('year', 'A'),
+])
+
+__tablename__ = 'pandasdatasources'
+type = 'pandas'
+baselink = 'pandasdatasourcemodelview' # url portion pointing to
ModelView endpoint
+column_class = PandasColumn
+metric_class = PandasMetric
+
+name = Column(String(100), nullable=False)
+source_url = Column(String(1000), nullable=False)
+format = Column(String(20), nullable=False)
+additional_parameters = Column(JSONType)
+
+user_id = Column(Integer, ForeignKey('ab_user.id'))
+owner = relationship(
+sm.user_model,
+
