Skip to content

Bug fixes #13

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
sibalm wants to merge 3 commits into
base: master
Choose a base branch
from
Open

Bug fixes #13

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
f0e1b70
Implement more compliant element redefinition in pandas dataframes
sibalm May 22, 2026
37a76f6
Restore correct indentation in loop
sibalm May 22, 2026
bea30c1
Fix patch length calculation by using resolution in range
sibalm May 22, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
157 changes: 78 additions & 79 deletions library/meso.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -158,7 +158,7 @@ def shear_group(rotation, sign, myfinaldata, az_shear, labels, resolution, dista
a=len(np.where(indices[1]==m)[0])
vec_width.append(distance[0,m]*a)
maxwidth=np.nanmax(vec_width)
maxlen=len(np.unique(indices[1]))
maxlen=len(np.unique(indices[1]))*resolution
ratio=maxlen/maxwidth
rankvel=(dvel-min_rvel)/(min_rvel)
rankvort=(vort-min_vort)/(4*min_vort)
Expand Down Expand Up @@ -216,10 +216,9 @@ def tower(rotation, areas, radar, shear, time, path):
for ID in n:
obj=prop.where(prop["v_ID"]==ID).dropna()
if len(obj)<1: continue
towers["ID"][ID]=ID
#towers["trtlat"][ID]=obj["trtlat"].values
#towers["trtlon"][ID]=obj["trtlon"].values
towers["radar"][ID]=np.unique(obj["radar"].values)
towers = towers.astype(object)
towers.loc[ID, "ID"]=ID
towers.at[ID, "radar"]=np.unique(obj["radar"].values)
r_range=[]
r_elev=[]
r_n=[]
Expand All @@ -229,30 +228,30 @@ def tower(rotation, areas, radar, shear, time, path):
r_n.append(len(o["vol"]))
r_elev.append(len(np.unique(o["elevation"])))
if n == 'A':
towers["A"][ID]=1
towers["A_range"][ID]=np.average(o["range"], weights=o["size"])*0.5
towers["A_n"][ID]=len(o["vol"])
towers["A_el"][ID]=len(np.unique(o["elevation"]))
towers.loc[ID, "A"]=1
towers.loc[ID, "A_range"]=np.average(o["range"], weights=o["size"])*0.5
towers.loc[ID, "A_n"]=len(o["vol"])
towers.loc[ID, "A_el"]=len(np.unique(o["elevation"]))
if n == 'D':
towers["D"][ID]=1
towers["D_range"][ID]=np.average(o["range"], weights=o["size"])*0.5
towers["D_n"][ID]=len(o["vol"])
towers["D_el"][ID]=len(np.unique(o["elevation"]))
towers.loc[ID, "D"]=1
towers.loc[ID, "D_range"]=np.average(o["range"], weights=o["size"])*0.5
towers.loc[ID, "D_n"]=len(o["vol"])
towers.loc[ID, "D_el"]=len(np.unique(o["elevation"]))
if n == 'L':
towers["L"][ID]=1
towers["L_range"][ID]=np.average(o["range"], weights=o["size"])*0.5
towers["L_n"][ID]=len(o["vol"])
towers["L_el"][ID]=len(np.unique(o["elevation"]))
towers.loc[ID, "L"]=1
towers.loc[ID, "L_range"]=np.average(o["range"], weights=o["size"])*0.5
towers.loc[ID, "L_n"]=len(o["vol"])
towers.loc[ID, "L_el"]=len(np.unique(o["elevation"]))
if n == 'P':
towers["P"][ID]=1
towers["P_range"][ID]=np.average(o["range"], weights=o["size"])*0.5
towers["P_n"][ID]=len(o["vol"])
towers["P_el"][ID]=len(np.unique(o["elevation"]))
towers.loc[ID, "P"]=1
towers.loc[ID, "P_range"]=np.average(o["range"], weights=o["size"])*0.5
towers.loc[ID, "P_n"]=len(o["vol"])
towers.loc[ID, "P_el"]=len(np.unique(o["elevation"]))
if n == 'W':
towers["W"][ID]=1
towers["W_range"][ID]=np.average(o["range"], weights=o["size"])*0.5
towers["W_n"][ID]=len(o["vol"])
towers["W_el"][ID]=len(np.unique(o["elevation"]))
towers.loc[ID, "W"]=1
towers.loc[ID, "W_range"]=np.average(o["range"], weights=o["size"])*0.5
towers.loc[ID, "W_n"]=len(o["vol"])
towers.loc[ID, "W_el"]=len(np.unique(o["elevation"]))
# Identify minimum range from any detecting radar
# Establish range-dependent depth threshold
ra=np.nanmin([towers.A_range[ID],towers.D_range[ID],towers.L_range[ID],towers.P_range[ID],towers.W_range[ID]])
Expand All @@ -264,70 +263,70 @@ def tower(rotation, areas, radar, shear, time, path):
dz_min=shear["zu"]-shear["zu"]*((20-ra)/20)
print("Minimum range, depth threshold", ra,dz_min)

towers["dz"][ID]=max(obj["z"])-min(obj["z"])
towers.loc[ID, "dz"]=max(obj["z"])-min(obj["z"])
# If depth threshold not met, discard 3D object
if towers["dz"][ID]<dz_min: towers.loc[ID]=np.nan; print('shear area too shallow', ID); continue

# All criteria met, fill rotation-tower dataframe with percentiles of 2D patches
towers["z_0"][ID]=np.nanmin(obj["z"])
towers["z_10"][ID]=np.percentile(obj["z"],10)
towers["z_25"][ID]=np.percentile(obj["z"],25)
towers["z_50"][ID]=np.percentile(obj["z"],50)
towers["z_75"][ID]=np.percentile(obj["z"],75)
towers["z_90"][ID]=np.percentile(obj["z"],90)
towers["z_100"][ID]=np.nanmax(obj["z"])
towers["z_IQR"][ID]=np.percentile(obj["z"],75)-np.percentile(obj["z"],25)
towers["z_mean"][ID]=np.nanmean(obj["z"])
towers.loc[ID, "z_0"]=np.nanmin(obj["z"])
towers.loc[ID, "z_10"]=np.percentile(obj["z"],10)
towers.loc[ID, "z_25"]=np.percentile(obj["z"],25)
towers.loc[ID, "z_50"]=np.percentile(obj["z"],50)
towers.loc[ID, "z_75"]=np.percentile(obj["z"],75)
towers.loc[ID, "z_90"]=np.percentile(obj["z"],90)
towers.loc[ID, "z_100"]=np.nanmax(obj["z"])
towers.loc[ID, "z_IQR"]=np.percentile(obj["z"],75)-np.percentile(obj["z"],25)
towers.loc[ID, "z_mean"]=np.nanmean(obj["z"])

towers["d_0"][ID]=np.nanmin(obj["diam"])
towers["d_10"][ID]=np.percentile(obj["diam"],10)
towers["d_25"][ID]=np.percentile(obj["diam"],25)
towers["d_50"][ID]=np.percentile(obj["diam"],50)
towers["d_75"][ID]=np.percentile(obj["diam"],75)
towers["d_90"][ID]=np.percentile(obj["diam"],90)
towers["d_100"][ID]=np.nanmax(obj["diam"])
towers["d_IQR"][ID]=np.percentile(obj["diam"],75)-np.percentile(obj["diam"],25)
towers["d_mean"][ID]=np.nanmean(obj["diam"])
towers.loc[ID, "d_0"]=np.nanmin(obj["diam"])
towers.loc[ID, "d_10"]=np.percentile(obj["diam"],10)
towers.loc[ID, "d_25"]=np.percentile(obj["diam"],25)
towers.loc[ID, "d_50"]=np.percentile(obj["diam"],50)
towers.loc[ID, "d_75"]=np.percentile(obj["diam"],75)
towers.loc[ID, "d_90"]=np.percentile(obj["diam"],90)
towers.loc[ID, "d_100"]=np.nanmax(obj["diam"])
towers.loc[ID, "d_IQR"]=np.percentile(obj["diam"],75)-np.percentile(obj["diam"],25)
towers.loc[ID, "d_mean"]=np.nanmean(obj["diam"])

towers["r_0"][ID]=np.nanmin(obj["dvel"])
towers["r_10"][ID]=np.percentile(obj["dvel"],10)
towers["r_25"][ID]=np.percentile(obj["dvel"],25)
towers["r_50"][ID]=np.percentile(obj["dvel"],50)
towers["r_75"][ID]=np.percentile(obj["dvel"],75)
towers["r_90"][ID]=np.percentile(obj["dvel"],90)
towers["r_100"][ID]=np.nanmax(obj["dvel"])
towers["r_IQR"][ID]=np.percentile(obj["dvel"],75)-np.percentile(obj["dvel"],25)
towers["r_mean"][ID]=np.nanmean(obj["dvel"])
towers.loc[ID, "r_0"]=np.nanmin(obj["dvel"])
towers.loc[ID, "r_10"]=np.percentile(obj["dvel"],10)
towers.loc[ID, "r_25"]=np.percentile(obj["dvel"],25)
towers.loc[ID, "r_50"]=np.percentile(obj["dvel"],50)
towers.loc[ID, "r_75"]=np.percentile(obj["dvel"],75)
towers.loc[ID, "r_90"]=np.percentile(obj["dvel"],90)
towers.loc[ID, "r_100"]=np.nanmax(obj["dvel"])
towers.loc[ID, "r_IQR"]=np.percentile(obj["dvel"],75)-np.percentile(obj["dvel"],25)
towers.loc[ID, "r_mean"]=np.nanmean(obj["dvel"])

towers["v_0"][ID]=np.nanmin(obj["vort"])
towers["v_10"][ID]=np.percentile(obj["vort"],10)
towers["v_25"][ID]=np.percentile(obj["vort"],25)
towers["v_50"][ID]=np.percentile(obj["vort"],50)
towers["v_75"][ID]=np.percentile(obj["vort"],75)
towers["v_90"][ID]=np.percentile(obj["vort"],90)
towers["v_100"][ID]=np.nanmax(obj["vort"])
towers["v_IQR"][ID]=np.percentile(obj["vort"],75)-np.percentile(obj["vort"],25)
towers["v_mean"][ID]=np.nanmean(obj["vort"])
towers.loc[ID, "v_0"]=np.nanmin(obj["vort"])
towers.loc[ID, "v_10"]=np.percentile(obj["vort"],10)
towers.loc[ID, "v_25"]=np.percentile(obj["vort"],25)
towers.loc[ID, "v_50"]=np.percentile(obj["vort"],50)
towers.loc[ID, "v_75"]=np.percentile(obj["vort"],75)
towers.loc[ID, "v_90"]=np.percentile(obj["vort"],90)
towers.loc[ID, "v_100"]=np.nanmax(obj["vort"])
towers.loc[ID, "v_IQR"]=np.percentile(obj["vort"],75)-np.percentile(obj["vort"],25)
towers.loc[ID, "v_mean"]=np.nanmean(obj["vort"])

towers["rank_0"][ID]=np.nanmin(obj["rank"])
towers["rank_10"][ID]=np.percentile(obj["rank"],10)
towers["rank_25"][ID]=np.percentile(obj["rank"],25)
towers["rank_50"][ID]=np.percentile(obj["rank"],50)
towers["rank_75"][ID]=np.percentile(obj["rank"],75)
towers["rank_90"][ID]=np.percentile(obj["rank"],90)
towers["rank_100"][ID]=np.nanmax(obj["rank"])
towers["rank_IQR"][ID]=np.percentile(obj["rank"],75)-np.percentile(obj["rank"],25)
towers["rank_mean"][ID]=np.nanmean(obj["rank"])
towers.loc[ID, "rank_0"]=np.nanmin(obj["rank"])
towers.loc[ID, "rank_10"]=np.percentile(obj["rank"],10)
towers.loc[ID, "rank_25"]=np.percentile(obj["rank"],25)
towers.loc[ID, "rank_50"]=np.percentile(obj["rank"],50)
towers.loc[ID, "rank_75"]=np.percentile(obj["rank"],75)
towers.loc[ID, "rank_90"]=np.percentile(obj["rank"],90)
towers.loc[ID, "rank_100"]=np.nanmax(obj["rank"])
towers.loc[ID, "rank_IQR"]=np.percentile(obj["rank"],75)-np.percentile(obj["rank"],25)
towers.loc[ID, "rank_mean"]=np.nanmean(obj["rank"])

towers["size_sum"][ID]=np.sum(obj["size"])
towers["size_mean"][ID]=np.nanmean(obj["size"])
towers["vol_sum"][ID]=np.sum(obj["vol"])
towers["vol_mean"][ID]=np.nanmean(obj["vol"])
towers.loc[ID, "size_sum"]=np.sum(obj["size"])
towers.loc[ID, "size_mean"]=np.nanmean(obj["size"])
towers.loc[ID, "vol_sum"]=np.sum(obj["vol"])
towers.loc[ID, "vol_mean"]=np.nanmean(obj["vol"])

towers["x"][ID]=np.average(obj["x"], weights=obj["size"])
towers["y"][ID]=np.average(obj["y"], weights=obj["size"])
towers["dz"][ID]=max(obj["z"])-min(obj["z"])
towers["time"][ID]=time
towers.loc[ID, "x"]=np.average(obj["x"], weights=obj["size"])
towers.loc[ID, "y"]=np.average(obj["y"], weights=obj["size"])
towers.loc[ID, "dz"]=max(obj["z"])-min(obj["z"])
towers.loc[ID, "time"]=time
print('Object merged; depth, rank, vorticity and rvel: ',towers["dz"][ID],towers["rank_90"][ID],towers["v_90"][ID],towers["r_90"][ID])
towers=towers.dropna()
print("Towers found: ", len(towers))
Expand Down
16 changes: 9 additions & 7 deletions realtime_parallel.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -242,13 +242,15 @@ def radel_processor (rotation_pos, rotation_neg, rels, radar, cartesian, path, s
ids=ids[ids>0]
# process each thunderstorm individually
for ii in ids:
rotation_pos1, rotation_neg1 = meso.cell_loop(ii, l_mask, az_shear, mfd_conv, rotation_pos1, rotation_neg1, distance, resolution, shear, radar, coord, timelist, r, el)
rotation_pos["prop"]=pd.concat([rotation_pos["prop"],rotation_pos1["prop"]], ignore_index=True)
rotation_neg["prop"]=pd.concat([rotation_neg["prop"],rotation_neg1["prop"]], ignore_index=True)
rotation_pos["shear_objects"].append(rotation_pos1["shear_objects"])
rotation_neg["shear_objects"].append(rotation_neg1["shear_objects"])
rotation_pos["shear_ID"].append(rotation_pos1["shear_ID"])
rotation_neg["shear_ID"].append(rotation_neg1["shear_ID"])
# dict values are lists that grow within the loop
rotation_pos1, rotation_neg1 = meso.cell_loop(ii, l_mask, az_shear, mfd_conv, rotation_pos1, rotation_neg1, distance, resolution, shear, radar, coord, timelist, r, el)

rotation_pos["prop"]=pd.concat([rotation_pos["prop"],rotation_pos1["prop"]], ignore_index=True)
rotation_neg["prop"]=pd.concat([rotation_neg["prop"],rotation_neg1["prop"]], ignore_index=True)
rotation_pos["shear_objects"].append(rotation_pos1["shear_objects"])
rotation_neg["shear_objects"].append(rotation_neg1["shear_objects"])
rotation_pos["shear_ID"].append(rotation_pos1["shear_ID"])
rotation_neg["shear_ID"].append(rotation_neg1["shear_ID"])

return_dict[el]= rotation_pos, rotation_neg

Expand Down