Hugging Face's logo

Spaces:
parsimony-dev
/
README
Running

App Files Community
ockhamsh commited on
Commit
92e9bc8
·
verified ·
1 Parent(s): 0f7df47

Simplify org card: plain markdown, no empty card boxes

Browse files
Files changed (1) hide show
  1. README.md +36 -64
README.md CHANGED
@@ -7,77 +7,49 @@ sdk: static
7
  pinned: false
8
  ---
9
 
10
- <div class="grid lg:grid-cols-3 gap-x-4 gap-y-7">
11
 
12
- <p class="lg:col-span-3">
13
- <strong>Pre-built searchable indexes of financial and economic data series for <a href="https://github.com/ockham-sh/parsimony" class="underline">parsimony</a>.</strong><br/>
14
- Each dataset here is a drop-in catalog — a parquet of series metadata plus a FAISS embedding index — that lets parsimony connectors resolve a natural-language query (<em>"Swiss policy rate"</em>, <em>"US 10-year yield"</em>) to a provider-native series ID in milliseconds, without hitting the upstream provider.
15
- </p>
16
 
17
- <a href="https://github.com/ockham-sh/parsimony" class="block overflow-hidden group">
18
- <div class="w-full h-40 object-cover mb-2 rounded-lg flex items-center justify-center bg-gray-50 border border-gray-200">
19
- <pre class="text-xs text-gray-800 whitespace-pre-wrap px-4">pip install parsimony \
20
- parsimony-sdmx \
21
- parsimony-fred</pre>
22
- </div>
23
- <div class="underline">parsimony on GitHub</div>
24
- </a>
25
 
26
- <a href="https://docs.parsimony.dev" class="block overflow-hidden group">
27
- <div class="w-full h-40 object-cover mb-2 rounded-lg flex items-center justify-center bg-gray-50 border border-gray-200">
28
- <span class="text-sm text-gray-700">docs.parsimony.dev</span>
29
- </div>
30
- <div class="underline">Documentation</div>
31
- </a>
32
 
33
- <a href="https://pypi.org/project/parsimony-core/" class="block overflow-hidden group">
34
- <div class="w-full h-40 object-cover mb-2 rounded-lg flex items-center justify-center bg-gray-50 border border-gray-200">
35
- <span class="text-sm text-gray-700">parsimony-core on PyPI</span>
36
- </div>
37
- <div class="underline">PyPI</div>
38
- </a>
39
 
40
- <div class="lg:col-span-3">
41
- <h3>What's in each catalog</h3>
42
- <ul>
43
- <li><code>entries.parquet</code> — one row per data series with provider ID, title, units, frequency, and provider-specific attributes.</li>
44
- <li><code>embeddings.faiss</code> — dense vector index over series titles and descriptions for semantic lookup.</li>
45
- <li><code>meta.json</code> — provider, snapshot timestamp, embedding model, and row counts.</li>
46
- </ul>
47
- <p>Catalogs are pulled on-demand by the matching connector (e.g. <code>parsimony-sdmx</code> reads every central-bank catalog below) and cached locally. No API keys required to read the indexes — keys are only needed when you fetch actual series data from the upstream provider.</p>
48
- </div>
49
 
50
- <div class="lg:col-span-3">
51
- <h3>Available catalogs</h3>
52
- <table>
53
- <thead><tr><th>Dataset</th><th>Provider</th><th>Protocol</th></tr></thead>
54
- <tbody>
55
- <tr><td><a href="https://huggingface.co/datasets/parsimony-dev/treasury">treasury</a></td><td>U.S. Department of the Treasury</td><td>Fiscal Data API</td></tr>
56
- <tr><td><a href="https://huggingface.co/datasets/parsimony-dev/snb">snb</a></td><td>Swiss National Bank</td><td>SDMX</td></tr>
57
- <tr><td><a href="https://huggingface.co/datasets/parsimony-dev/riksbank">riksbank</a></td><td>Sveriges Riksbank</td><td>SDMX</td></tr>
58
- <tr><td><a href="https://huggingface.co/datasets/parsimony-dev/boc">boc</a></td><td>Bank of Canada</td><td>SDMX</td></tr>
59
- <tr><td><a href="https://huggingface.co/datasets/parsimony-dev/rba">rba</a></td><td>Reserve Bank of Australia</td><td>SDMX</td></tr>
60
- <tr><td><a href="https://huggingface.co/datasets/parsimony-dev/bde">bde</a></td><td>Banco de España</td><td>SDMX</td></tr>
61
- </tbody>
62
- </table>
63
- <p>More providers land as their connectors graduate. The publishing pipeline is open — see <a href="https://github.com/ockham-sh/parsimony-connectors">parsimony-connectors</a> for how catalogs are generated and contributed.</p>
64
- </div>
65
 
66
- <div class="lg:col-span-3">
67
- <h3>Quick start</h3>
68
- <pre class="text-xs"><code>from parsimony import Parsimony
 
 
 
 
 
69
 
70
- p = Parsimony()
71
- # Resolves "swiss policy rate" against the snb catalog, then fetches from SNB.
72
- series = p.find("swiss policy rate").fetch()
73
- print(series.tail())
74
- </code></pre>
75
- </div>
 
 
 
 
 
76
 
77
- <div class="lg:col-span-3">
78
- <p class="text-sm text-gray-600">
79
- Maintained by <a href="https://github.com/ockham-sh">Ockham</a>. Issues and contributions: <a href="https://github.com/ockham-sh/parsimony-connectors">parsimony-connectors</a>.
80
- </p>
81
- </div>
82
 
83
- </div>
 
7
  pinned: false
8
  ---
9
 
10
+ # parsimony
11
 
12
+ **Pre-built searchable indexes of financial and economic data series.**
 
 
 
13
 
14
+ Each dataset here is a drop-in catalog — a parquet of series metadata plus a FAISS embedding index — that lets [parsimony](https://github.com/ockham-sh/parsimony) resolve a natural-language query like *"swiss policy rate"* or *"US 10-year yield"* to a provider-native series ID in milliseconds, without hitting the upstream provider.
 
 
 
 
 
 
 
15
 
16
+ ```bash
17
+ pip install parsimony parsimony-sdmx
18
+ ```
 
 
 
19
 
20
+ ```python
21
+ from parsimony import Parsimony
 
 
 
 
22
 
23
+ p = Parsimony()
24
+ series = p.find("swiss policy rate").fetch()
25
+ ```
 
 
 
 
 
 
26
 
27
+ ## Catalogs
 
 
 
 
 
 
 
 
 
 
 
 
 
 
28
 
29
+ | Dataset | Provider | Protocol |
30
+ |---|---|---|
31
+ | [`treasury`](https://huggingface.co/datasets/parsimony-dev/treasury) | U.S. Department of the Treasury | Fiscal Data |
32
+ | [`snb`](https://huggingface.co/datasets/parsimony-dev/snb) | Swiss National Bank | SDMX |
33
+ | [`riksbank`](https://huggingface.co/datasets/parsimony-dev/riksbank) | Sveriges Riksbank | SDMX |
34
+ | [`boc`](https://huggingface.co/datasets/parsimony-dev/boc) | Bank of Canada | SDMX |
35
+ | [`rba`](https://huggingface.co/datasets/parsimony-dev/rba) | Reserve Bank of Australia | SDMX |
36
+ | [`bde`](https://huggingface.co/datasets/parsimony-dev/bde) | Banco de España | SDMX |
37
 
38
+ More providers land as their connectors graduate.
39
+
40
+ ## Anatomy of a catalog
41
+
42
+ - `entries.parquet` — one row per series with provider ID, title, units, frequency.
43
+ - `embeddings.faiss` — dense vector index over titles and descriptions for semantic lookup.
44
+ - `meta.json` — provider, snapshot timestamp, embedding model, row counts.
45
+
46
+ Catalogs are pulled on-demand by the matching connector and cached locally. No API keys required to read the indexes — keys are only needed when you fetch the actual series from the upstream provider.
47
+
48
+ ## Links
49
 
50
+ - **Kernel** — [github.com/ockham-sh/parsimony](https://github.com/ockham-sh/parsimony)
51
+ - **Connectors** — [github.com/ockham-sh/parsimony-connectors](https://github.com/ockham-sh/parsimony-connectors)
52
+ - **Docs** — [docs.parsimony.dev](https://docs.parsimony.dev)
53
+ - **PyPI** — [`parsimony-core`](https://pypi.org/project/parsimony-core/)
 
54
 
55
+ Maintained by [Ockham](https://github.com/ockham-sh) · Apache 2.0