Skip to content

Commit

Permalink
Update Documentation.tsx
Browse files Browse the repository at this point in the history
  • Loading branch information
@tomohiro-sawada
tomohiro-sawada authored Oct 12, 2023
1 parent 0903a13 commit 0043bdb
Showing 1 changed file with 153 additions and 56 deletions.
209 changes: 153 additions & 56 deletions src/screens/Documentation/Documentation.tsx
View file
Original file line number Diff line number Diff line change
Expand Up @@ -57,177 +57,275 @@ const Documentation: React.FC = () => {
<Text mb={4}>
We provide a simple API to access the Advanced Reasoning Benchmark (ARB). The API currently supports standard HTTP GET requests.
</Text>
{/* <Heading as="h2" size="xl" mb={6}>

{/* <Heading as="h2" size="xl" mb={6}>
API Calls
</Heading>
<Text mb={4}>
We have three different types of API calls for retrieving problems. See <Link href="https://app.swaggerhub.com/apis-docs/arb-dataset/arb-api/1.0.5" color="teal.500">here</Link> and select ARB API server as the mock server to see what the outputs look like.
</Text> */}





<Heading as="h3" size="lg" mb={4}>
Get problem by category:
Categories without Images.
</Heading>
<Text mb={4}>
You can retrieve all the problems in a particular subject area by the
following API call. Acceptable values for `category` are `law`,
`math`, `mcatReading`, etc.
For law, problems are retrieved according to the following:
</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
import requests
<br />
response =
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/law/&quot;)
<br />
data = response.json()
</Code>
</Box>

<Text mb={4}>
For math, you can retrieve the problems according to the following template:
</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>https://advanced-reasoning-benchmark.netlify.app/api/lib/{"{category}"}</Code>
<Code>
https://advanced-reasoning-benchmark.netlify.app/api/lib/math/{"{answer-type}"}/
</Code>
</Box>
<Text fontWeight="bold" mb={2}>
Math Numerical:
<Text mb={4}>
where acceptable values for `answer-type` are `numerical`, `symbolic`, and `prooflike`.
</Text>
<Text mb={4}>
Numerical
</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
import requests
<br />
response =
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/math&quot;)
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/math/numerical&quot;)
<br />
data = response.json()
</Code>
</Box>

<Text fontWeight="bold" mb={2}>
Math Symbolic:
<Text mb={4}>
Symbolic
</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
import requests
<br />
response =
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/testSplit/math/symbolic&quot;)
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/math/symbolic&quot;)
<br />
data = response.json()
</Code>
</Box>

<Text fontWeight="bold" mb={2}>
Math Proofs:
<Text mb={4}>
Proof-like
</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
import requests
<br />
response =
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/testSplit/math/proof&quot;)
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/math/prooflike&quot;)
<br />
data = response.json()
</Code>
</Box>

<Text fontWeight="bold" mb={2}>
Physics Numerical:

<Text mb={4}>
For MCAT Reading, you can retrieve the problems according to the following template:
</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
https://advanced-reasoning-benchmark.netlify.app/api/lib/mcatReading/{"{split}"}
</Code>
</Box>
<Text mb={4}>
where acceptable values for `split` are `val` and `test`.
</Text>
<Text mb={4}>
MCAT Reading Validation Split:
</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
import requests
<br />
response =
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/physics/val&quot;)
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/mcatReading/val&quot;)
<br />
data = response.json()
</Code>
</Box>

<Text fontWeight="bold" mb={2}>
Physics Symbolic:
</Text>
<Text mb={4}>
MCAT Reading Test Split:
</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
import requests
<br />
response =
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/testSplit/physics/val&quot;)
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/mcatReading/test&quot;)
<br />
data = response.json()
</Code>
</Box>

<Text fontWeight="bold" mb={2}>
Law:






<Heading as="h3" size="lg" mb={4}>
Categories without Images.
</Heading>
<Text mb={4}>
For physics and MCAT Science, you can specify whether you want the problems with or without images.

</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
https://advanced-reasoning-benchmark.netlify.app/api/lib/{"{category}"}/{"{answer-type}"}/{"{modality}"}
</Code>
</Box>
<Text mb={4}>
For physics, acceptable values for `{"{answer-type}"}` are `numerical` and `symbolic`. Acceptable values for `{"{modality}"}` are `img` and `noimg`.
</Text>
<Text mb={4}>
Physics Numerical without Images
</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
import requests
<br />
response =
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/law&quot;)
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/physics/numerical/noimg&quot;)
<br />
data = response.json()
</Code>
</Box>

<Text fontWeight="bold" mb={2}>
MCAT Reading:
<Text mb={4}>
Physics Numerical with Images
</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
import requests
<br />
response =
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/mcatReading&quot;)
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/physics/numerical/img&quot;)
<br />
data = response.json()
</Code>
</Box>

<Heading as="h3" size="lg" mb={4}>
Get a specific problem by id within category
</Heading>
<Text mb={4}>
Physics Symbolic without Images
</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
https://advanced-reasoning-benchmark.netlify.app/api/lib/{"{category}"}/{"{id}"}
import requests
<br />
response =
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/physics/symbolic/noimg&quot;)
<br />
data = response.json()
</Code>
</Box>
<Text mb={4}>
Physics Symbolic with Images
</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
import requests
<br />
response =
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/math/1234&quot;)
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/physics/symbolic/img&quot;)
<br />
problem = response.json()
data = response.json()
</Code>
</Box>

<Heading as="h3" size="lg" mb={4}>
Get problem by category variation (e.g. img vs val)
</Heading>
<Text mb={4}>
For physics and MCAT Science, you can specify whether you want the problems with or without images.

For MCAT Science, `{"{answer-type}"}` should be filled in with the splits, namely `val` and `test`. Acceptable values for `{"{modality}"}` are `img` and `noimg`. For example, if you want the MCAT Science from the validation split without images, you can use the following template:
</Text>
<Text mb={4}>
MCAT Science Validation Split without Image
</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
https://advanced-reasoning-benchmark.netlify.app/api/lib/{"{category}"}/{"{variation}"}
import requests
<br />
response =
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/mcatScience/val/noimg&quot;)
<br />
data = response.json()
</Code>
</Box>
<Text fontWeight="bold" mb={2}>
Physics Numerical w/ Images:
</Box>
<Text mb={4}>
MCAT Science Validation Split with Image
</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
import requests
<br />
response =
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/physics_numerical/img&quot;)
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/mcatScience/val/img&quot;)
<br />
variation_data = response.json()
data = response.json()
</Code>
</Box>
<Text fontWeight="bold" mb={2}>
Physics Symbolic w/ Images:
</Box>
<Text mb={4}>
MCAT Science Test Split without Image
</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
import requests
<br />
response =
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/testSplit/physics/img&quot;)
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/mcatScience/test/noimg&quot;)
<br />
variation_data = response.json()
data = response.json()
</Code>
</Box>
<Text mb={4}>
MCAT Science Test Split with Image
</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
import requests
<br />
response =
requests.get(&quot;https://advanced-reasoning-benchmark.netlify.app/api/lib/mcatScience/test/img&quot;)
<br />
data = response.json()
</Code>
</Box>


<Heading as="h3" size="lg" mb={4}>
Get a specific problem by id within category
</Heading>
<Text mb={4}>
For any of the above, if you terminate the API call with `/{"id"}`, then you can get a specific problem with that ID.

Some examples:

</Text>
<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
https://advanced-reasoning-benchmark.netlify.app/api/lib/mcatScience/test/noimg/64cdbc1978a46c58407af688
</Code>
</Box>


<Box p={4} mb={4} borderWidth="1px" borderRadius="md" bg="gray.50">
<Code>
https://advanced-reasoning-benchmark.netlify.app/api/lib/math/numerical/64ade9c30b1afac21d212df7
</Code>
</Box>
</Container>
Expand Down Expand Up @@ -256,4 +354,3 @@ const Documentation: React.FC = () => {
};

export default Documentation;

0 comments on commit 0043bdb

Please sign in to comment.