The seurat converted into anndata to analyze the trajectory inference occurred the weird circumstance

Hi ,
In my case, the seurat object using the sceasy algorithm to transfer into anndata object for trajectory inference analysis.
The code lying below:
import numpy as np
import pandas as pd
import matplotlib.pyplot as pl
from matplotlib import rcParams
import scanpy as sc
sc.pp.recipe_zheng17(adata)
sc.tl.pca(adata, svd_solver='arpack')
sc.pp.neighbors(adata, n_neighbors=4, n_pcs=20)
sc.tl.draw_graph(adata)
sc.pl.draw_graph(adata, color='paul15_clusters', legend_loc='on data')
The picture showing confused result posted below:

The object information:

adata
AnnData object with n_obs × n_vars = 17885 × 999
obs: 'orig.ident', 'nCount_RNA', 'nFeature_RNA', 'percent.mt', 'RNA_snn_res.0.5', 'seurat_clusters', 'pANN_0.25_0.02_752', 'DF.classifications_0.25_0.02_752', 'percent.rp', 'pANN_0.25_0.02_826', 'DF.classifications_0.25_0.02_826', 'group', 'celltype', 'n_counts_all'
var: 'vst.mean', 'vst.variance', 'vst.variance.expected', 'vst.variance.standardized', 'vst.variable', 'n_counts', 'mean', 'std'
uns: 'seurat_clusters_colors', 'log1p', 'pca', 'neighbors', 'draw_graph'
obsm: 'X_pca', 'X_tsne', 'X_umap', 'X_draw_graph_fr'
varm: 'PCs'
obsp: 'distances', 'connectivities'

Start by reproducing the supplied AnnData conversion and preprocessing sequence with the listed 17,885-cell object. Inspect the converted metadata, PCA, neighbor graph, and draw_graph outputs to determine which step produces the unexpected result. Done means identifying an actionable cause or confirming the required conversion or analysis settings.