总览 · ← 返回运行 20261002-202908-search-t1-scr-C

节点 n46

改了什么

方法说明?节点程序自带的 METHOD.md：这个程序做了什么、为什么。

来自 programs.git b824430772:solution/METHOD.md

调研员的计划

代码改动?这个节点的程序和父节点程序的逐行差别：绿色是新增，红色是删除。

对比：这个提交的上一版（种子程序：相对空仓库）。改动的文件：solution/EXECUTION.json +1 −0、solution/METHOD.md +14 −0、solution/README.md +4 −0、solution/g37_common.py +101 −0、solution/run.py +120 −0

diff --git a/solution/EXECUTION.json b/solution/EXECUTION.jsonnew file mode 100644index 0000000..9d5125c--- /dev/null+++ b/solution/EXECUTION.json@@ -0,0 +1 @@+{"gpu": false}diff --git a/solution/METHOD.md b/solution/METHOD.mdnew file mode 100644index 0000000..30c3328--- /dev/null+++ b/solution/METHOD.md@@ -0,0 +1,14 @@+## 改了什么+从父节点 ot_moscot（OT 耦合 + 位移外推）完全转向 lowrank_shape 方向：+1. 去掉 moscot TemporalProblem / JAX / torch 依赖，改用 sklearn LedoitWolf 收缩协方差估计。+2. 对每个同时出现在两个输入阶段的细胞类型，在 PCA 空间分别估计两阶段的收缩协方差，计算形状变换 T = S_last^{1/2} @ S_prev^{-1/2}，向单位阵收缩（ALPHA=0.3），作用于 last 阶段的中心化残差。均值不移动，只改变群体形状（协变结构）。+3. 解码回基因空间时只加在非零条目上（addnz），保持稀疏模式，夹到 >=0。+4. 保留生长率加权重抽样（WOT birth-death）决定输出组成。+5. 机制对照：LOWRANK_SHAPE_OFF=1 时 T=I（形状不变），提交时保持打开。+6. 单输入阶段退化为生长加权重抽样复制。+理由：父节点 direction 最弱（49.14），但本方向不针对 direction，而是针对 covariation（父节点 52.52，有提升空间）。通过改变群体形状而非均值，预期在 covariation 上获益而不损害 cell_state / de_recovery。+## 用到的知识与出处+- Ledoit-Wolf 收缩协方差估计：scikit-learn LedoitWolf 文档；Ledoit & Wolf (2004) "A well-conditioned estimator for large-dimensional covariance matrices", J. Multivariate Anal.+- 协方差外推（T = S_last^{1/2} @ S_prev^{-1/2}）是自研假设，不是该估计器已验证的功能（方向卡已注明）。+- 生长率先验：moscot 内置小鼠增殖/凋亡基因列表（Schiebinger 2019, Cell; 通用基因功能注释，非禁窗测量）。+- PCA + HVG 预处理：g37_common.embed（父节点已有）。diff --git a/solution/README.md b/solution/README.mdnew file mode 100644index 0000000..dc1cbaf--- /dev/null+++ b/solution/README.md@@ -0,0 +1,4 @@+# ot_moscot++Waddington-OT / moscot：两个最新输入阶段在联合 PCA 上做非平衡熵 OT 耦合（WOT 增殖/凋亡先验生长），最新阶段按 g^Δt 重抽样，每个细胞沿“kNN 平滑位置 − 耦合祖先均值”再走一个等比例步长，只加在非零基因上，夹到 ≥0。+来自 G37 候选 modeling/candidates/T1/ot_moscot；addnz 解码是在 X3 上选的（详见 METHOD.md）。纯 CPU，final 约 70 s、6.7 GB。diff --git a/solution/g37_common.py b/solution/g37_common.pynew file mode 100644index 0000000..8c991b3--- /dev/null+++ b/solution/g37_common.py@@ -0,0 +1,101 @@+"""Helpers of the ot_moscot seed (copy of modeling/candidates/T1/ot_moscot/g37_common.py, G37; stage_pair reads every+input stage of the view).++- ``stage_pair``: the two latest input stages (official, or external in proxy2 / test views) and the gene mask on+  which both are really measured (external stages lack part of the panel; filled columns must not drive dynamics);+- ``embed``: HVG -> z-score (clip 10) -> PCA, fitted on the input stages only (scanpy-style preprocessing);+- ``knn_mean``: per-cell kNN average in the embedding (smooths single-cell noise before taking displacements);+- ``growth_rates``: Waddington-OT / moscot prior growth from proliferation and apoptosis gene scores.+"""++from __future__ import annotations++import numpy as np+from scipy import sparse++from src.task1_temporal.view_io import covered_mask, inputs_by_time, read_stage+++def stage_pair(view, manifest: dict, genes: list[str]):+    """(prev, last, mask, dt_in) for the two latest inputs; prev None with a single input stage."""+    stages = inputs_by_time(manifest, include_external=True)  # every input stage, whatever its source+    last_e = stages[-1]+    last = read_stage(view, last_e, genes)+    mask = covered_mask(view, last_e, genes)+    if len(stages) < 2:+        return None, last, mask, None, last_e+    prev_e = stages[-2]+    prev = read_stage(view, prev_e, genes)+    mask &= covered_mask(view, prev_e, genes)+    return prev, last, mask, float(last_e["time"] - prev_e["time"]), last_e+++def _gene_moments(X: sparse.csr_matrix):+    n = X.shape[0]+    mean = np.asarray(X.mean(axis=0)).ravel()+    sq = np.asarray(X.multiply(X).sum(axis=0)).ravel() / n+    return mean, np.maximum(sq - mean ** 2, 0.0)+++def embed(mats: list, mask: np.ndarray, n_hvg: int, n_pcs: int, seed: int):+    """PCA coordinates of each matrix in ``mats`` (same order), fitted on all of them together.++    Returns (list of Z, info) with info = {hvg, mean, std, components} so displacements can be decoded."""+    from sklearn.decomposition import PCA++    allX = sparse.vstack(mats).tocsr()+    mean, var = _gene_moments(allX)+    var = np.where(mask, var, -1.0)+    hvg = np.sort(np.argsort(var)[::-1][:n_hvg])+    sub = allX[:, hvg].toarray().astype(np.float32)+    mu, sd = mean[hvg].astype(np.float32), np.sqrt(var[hvg]).astype(np.float32)+    sd[sd == 0] = 1.0+    sub -= mu+    sub /= sd+    np.clip(sub, -10, 10, out=sub)+    pca = PCA(n_components=n_pcs, svd_solver="randomized", random_state=seed)+    Z = pca.fit_transform(sub).astype(np.float32)+    out, start = [], 0+    for M in mats:+        out.append(Z[start:start + M.shape[0]])+        start += M.shape[0]+    return out, {"hvg": hvg, "mu": mu, "sd": sd, "components": pca.components_.astype(np.float32)}+++def knn_mean(Z: np.ndarray, X: sparse.csr_matrix, rows: np.ndarray, k: int) -> np.ndarray:+    """Dense (len(rows), n_genes): mean expression of the k nearest cells (in Z, the cell itself included)."""+    from sklearn.neighbors import NearestNeighbors++    nn = NearestNeighbors(n_neighbors=k).fit(Z)+    idx = nn.kneighbors(Z[rows], return_distance=False)+    W = sparse.csr_matrix((np.full(idx.size, 1.0 / k, dtype=np.float32), idx.ravel(),+                           np.arange(0, idx.size + 1, k)), shape=(len(rows), Z.shape[0]))+    return np.asarray((W @ X).todense(), dtype=np.float32)+++# Waddington-OT prior growth (Schiebinger et al. 2019, Cell; as implemented in moscot.base.problems.birth_death):+# birth = generalised logistic of the proliferation score, death = of the apoptosis score, g = exp(birth - death) / day.+def _gen_logistic(p, sup, inf, center, width):+    return inf + (sup - inf) / (1 + np.exp(-(p - center) / width))+++def growth_rates(adata_list: list, genes: list[str], mask: np.ndarray, seed: int) -> list[np.ndarray]:+    """Per-day growth rate of every cell of each AnnData (scores computed on all of them jointly)."""+    import anndata as ad+    import scanpy as sc+    from moscot.utils.data import apoptosis_markers, proliferation_markers++    ok = set(np.asarray(genes)[mask])+    joint = ad.concat(adata_list, join="inner")+    prolif = [g for g in proliferation_markers("mouse") if g in ok]+    apopt = [g for g in apoptosis_markers("mouse") if g in ok]+    sc.tl.score_genes(joint, prolif, score_name="proliferation", random_state=seed)+    sc.tl.score_genes(joint, apopt, score_name="apoptosis", random_state=seed)+    birth = _gen_logistic(joint.obs["proliferation"].to_numpy(float), 1.7, 0.3, 0.25, 0.5)+    death = _gen_logistic(joint.obs["apoptosis"].to_numpy(float), 1.7, 0.3, 0.1, 0.2)+    g = np.exp(birth - death)+    out, start = [], 0+    for a in adata_list:+        out.append(g[start:start + a.n_obs])+        start += a.n_obs+    return outdiff --git a/solution/run.py b/solution/run.pynew file mode 100644index 0000000..7f3f99b--- /dev/null+++ b/solution/run.py@@ -0,0 +1,120 @@+#!/usr/bin/env python3+"""lowrank_shape: per-type Ledoit-Wolf shrunk covariance evolution in PCA space.++For each cell type present in both input stages, estimate the shrunk covariance+of centered residuals in PCA space at both stages, compute the affine shape+transform T = S_last^{1/2} @ S_prev^{-1/2}, shrink towards identity by ALPHA,+and apply to last-stage centered residuals. Per-type means are not moved; only+the population shape (covariation structure) evolves.++One input stage: growth-weighted copy of the latest stage (no shape transform).+Control: LOWRANK_SHAPE_OFF=1 sets T = I (no shape change).+"""++from __future__ import annotations++import argparse+import os+import sys+from pathlib import Path++import numpy as np++sys.path.insert(0, str(Path(__file__).resolve().parent))+from g37_common import embed, growth_rates, stage_pair  # noqa: E402++from src.task1_temporal.view_io import load_manifest, panel_genes, target_n_cells, write_prediction  # noqa: E402++N_HVG = 2000+N_PCS = 30+RIDGE = 1e-2+ALPHA = 0.3+MIN_TYPE_CELLS = 15+MECHANISM_OFF = os.environ.get("LOWRANK_SHAPE_OFF", "0") == "1"+++def weighted_rows(w: np.ndarray, n: int, rng: np.random.Generator) -> np.ndarray:+    p = w / w.sum()+    return np.sort(rng.choice(len(w), size=n, replace=n > len(w), p=p))+++def compute_transform(R_prev: np.ndarray, R_last: np.ndarray) -> np.ndarray:+    from sklearn.covariance import LedoitWolf++    n_pcs = R_prev.shape[1]+    if MECHANISM_OFF:+        return np.eye(n_pcs, dtype=np.float32)+    lw_p = LedoitWolf().fit(R_prev)+    lw_l = LedoitWolf().fit(R_last)+    S_p = lw_p.covariance_ + RIDGE * np.eye(n_pcs)+    S_l = lw_l.covariance_ + RIDGE * np.eye(n_pcs)+    w_p, V_p = np.linalg.eigh(S_p)+    w_p = np.maximum(w_p, 1e-8)+    inv_sqrt_p = (V_p / np.sqrt(w_p)) @ V_p.T+    w_l, V_l = np.linalg.eigh(S_l)+    w_l = np.maximum(w_l, 0.0)+    sqrt_l = (V_l * np.sqrt(w_l)) @ V_l.T+    T = sqrt_l @ inv_sqrt_p+    return ((1.0 - ALPHA) * np.eye(n_pcs) + ALPHA * T).astype(np.float32)+++def main() -> None:+    parser = argparse.ArgumentParser()+    parser.add_argument("--data", required=True)+    parser.add_argument("--out", required=True)+    parser.add_argument("--seed", type=int, default=0)+    args = parser.parse_args()++    manifest = load_manifest(args.data)+    genes = panel_genes(args.data, manifest)+    prev, last, mask, dt_in, last_e = stage_pair(args.data, manifest, genes)+    dt_out = float(manifest["target"]["time"] - last_e["time"])+    rng = np.random.default_rng(args.seed)+    n = target_n_cells(manifest, last.n_obs)++    if prev is None:+        (g_last,) = growth_rates([last], genes, mask, args.seed)+        rows = weighted_rows(g_last ** dt_out, n, rng)+        write_prediction(last.X[rows], genes, args.out, seed=args.seed)+        return++    (Zp, Zl), info = embed([prev.X, last.X], mask, N_HVG, N_PCS, args.seed)+    g_prev, g_last = growth_rates([prev, last], genes, mask, args.seed)+    rows = weighted_rows(g_last ** dt_out, n, rng)++    ct_prev = prev.obs["celltype"].values+    ct_last = last.obs["celltype"].values+    common = set(ct_prev) & set(ct_last)++    Z_out = Zl[rows].copy()+    row_types = ct_last[rows]++    for ct in sorted(common):+        idx_p = np.where(ct_prev == ct)[0]+        idx_l = np.where(ct_last == ct)[0]+        if len(idx_p) < MIN_TYPE_CELLS or len(idx_l) < MIN_TYPE_CELLS:+            continue+        mean_p = Zp[idx_p].mean(axis=0)+        mean_l = Zl[idx_l].mean(axis=0)+        T = compute_transform(Zp[idx_p] - mean_p, Zl[idx_l] - mean_l)+        sel = np.where(row_types == ct)[0]+        if len(sel) == 0:+            continue+        centered = Z_out[sel] - mean_l+        Z_out[sel] = mean_l + centered @ T.T++    hvg = info["hvg"]+    V = info["components"]+    sd = info["sd"]+    delta_Z = Z_out - Zl[rows]+    delta_gene = (delta_Z @ V) * sd++    X = last.X[rows].toarray()+    nz = X[:, hvg] > 0+    X[:, hvg] += delta_gene * nz+    np.maximum(X, 0.0, out=X)+    write_prediction(X, genes, args.out, seed=args.seed)+++if __name__ == "__main__":+    main()

调研来源?调研员查到并用到的知识条目和文献检索结果（只列标题和编号）。

没有记录调研来源。

分析结果?分析员写的 ANALYSIS.json：改了什么、各组分数怎么变、假设是否成立、经验和下一步建议。

对话摘要?每个角色和大模型对话的统计：轮数、工具调用、用时、token 数和最后的回答摘录；原始记录只给路径。

只给统计和最后回答的摘录；完整对话请到原始记录位置里列出的文件看。

分析员

ANALYSIS.json written

native

审查员

REVIEW.json written

stream

<tool_call>
{"name": "read_file", "arguments": {"path": "/home/spark-longxinyang/.local/state/vec-report/calls/2-202908-search-t1-scr-C-n46-e131ce/work/PROMPT.md", "offset": 667}}
</tool_call>