Yes. Please read the Pandas Tutorial.
Please update your
frictionless version. Nowadays, Frictionless treats all JSON and YAML files as metadata by default so this error should not present.
Yes, you just need to create a Table Schema and re-use it (the same can be done for a Dialect):
If you work with a data package you can have
package.resources.schema = "path/to/schema.json" although it's not possible to share a subset of fields only the whole schema. For sharing a subset of fields you need to copy them.
Yes, your resource can have
resource.layout.header = False in Python.
Yes, you need to use
resource.layout.header_rows = [1,2,3] in Python or
--header-rows 1,2,3 in CLI. Also, Here is an example of a More Complex Use Case.
The main difference is that Google datasets search use schema.org, which supports JSON-LD. We currently don't support linked data like JSON-LD. But they are compatible. Something like DataCite is more specific than Frictionless. Frictionless is general by design, and can be expanded to be compatible with other schemas and standards.
Currently, Parquet data is not supported but it's on our roadmap.
Yes, please follow the Describe Guide but instead of local CSV path provide a SQLite url or Pandas dataframe object. For more information about individual data formats please take a look at Formats Tutorials.
We use JSON Schema to validate our metadata. JSON Schema (we call it profile) is a sort of meta-meta-data for us. We validate our metadata using JSON Schema (profiles).
If you use JSON Schema to describe your tabular data, you can use
Schema.from_jsonschema helper function to translate it to Table Schema.
Frictionless works with data sources that have tabular structure. Individual files in Frictionless can be
arrays so you can maintain nested data using these data types. On the other hand, you might need to use other tools like JSON Schema if structure of your data is to complex.
Yes, you can use
package.resources.schema.foreign_keys in Python to add foreign keys to your schema, and when you validate your data. If there are integrity errors, you will get a Foreign Key Error.
If the spec doesn't have (yet) this attribute, you can add it. For an examaple, please take a look at Describe Guide
We're planning to add this functionality to the CLI as well.
It's possible using Python:
If you're using zsh linux terminal instead of bash, is good to know that zsh uses square brackets for globbing / pattern matching
That means that if you need to pass literal square brackets as an argument to a command, you either need to escape them or quote the argument like this:
So, I this case: