Skip to content

Commit

Permalink
check if Metric exists in smartUnits
Browse files Browse the repository at this point in the history 
Signed-off-by: JoseSantosAMD <Jose.Santos@amd.com>
  • Loading branch information
@JoseSantosAMD
JoseSantosAMD committed Aug 25, 2023
1 parent c5aa3d3 commit 75c7b25
Showing 1 changed file with 115 additions and 114 deletions.
229 changes: 115 additions & 114 deletions src/omniperf_analyze/utils/tty.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -36,120 +36,121 @@
def smartUnits(df):
for idx, row in df[df["Unit"] == "Gb/s"].items():
for curr_metric in row:
curr_row = df[df["Metric"] == curr_metric]
if not curr_row.empty:
# fix values
if "Value" in curr_row:
vals = curr_row["Value"].values
new_units = []
percent_diff = 0

# if baseline
if isinstance(vals[0], np.ndarray):
val_1 = curr_row["Value"].values[0][0]
baseline = curr_row["Value"].values[0][1].split()
val_2 = float(baseline[0])
percent_diff = baseline[1]
vals = np.array([val_1, val_2])

# calculate units
for val in vals:
if isinstance(
val,
float,
):
if val < 0.001:
new_units.append("Kb/s")
elif val < 1:
new_units.append("Mb/s")
if len(new_units) == 2:
if new_units[0] == "Kb/s":
new_units[0] = "Mb/s"
else:
new_units.append("Gb/s")
if len(new_units) == 2:
new_units[0] = "Gb/s"
# Convert to new_units
if new_units[0] == "Mb/s":
vals = 1000 * vals
elif new_units[0] == "Kb/s":
vals = 1000000 * vals
vals = vals.tolist()
# if baseline
if len(new_units) == 2:
vals[1] = str(vals[1]) + " " + str(percent_diff)

df.loc[df["Metric"] == curr_metric, "Value"] = vals
df.loc[df["Metric"] == curr_metric, "Unit"] = new_units[0]

elif "Avg" in curr_row:
avg_vals = curr_row["Avg"].values
max_vals = curr_row["Max"].values
min_vals = curr_row["Min"].values
new_units = []
percent_diff = 0

# if baseline
if isinstance(avg_vals[0], np.ndarray):
avg_baseline = curr_row["Avg"].values[0][1].split()
avg_percent_diff = avg_baseline[1]
avg_val_1 = curr_row["Avg"].values[0][0]
avg_val_2 = float(avg_baseline[0])
avg_vals = np.array([avg_val_1, avg_val_2])

min_baseline = curr_row["Min"].values[0][1].split()
min_percent_diff = min_baseline[1]
min_val_1 = curr_row["Min"].values[0][0]
min_val_2 = float(min_baseline[0])
min_vals = np.array([min_val_1, min_val_2])

max_baseline = curr_row["Max"].values[0][1].split()
max_percent_diff = max_baseline[1]
max_val_1 = curr_row["Max"].values[0][0]
max_val_2 = float(max_baseline[0])
max_vals = np.array([max_val_1, max_val_2])

# calculate units
for val in avg_vals:
if isinstance(
val,
float,
):
if val < 0.001:
new_units.append("Kb/s")
elif val < 1:
new_units.append("Mb/s")
if len(new_units) == 2:
if new_units[0] == "Kb/s":
new_units[0] = "Mb/s"
else:
new_units.append("Gb/s")
if len(new_units) == 2:
new_units[0] = "Gb/s"

# Convert to new_units
if new_units[0] == "Mb/s":
avg_vals = 1000 * avg_vals
max_vals = 1000 * max_vals
min_vals = 1000 * min_vals
elif new_units[0] == "Kb/s":
avg_vals = 1000000 * avg_vals
max_vals = 1000000 * max_vals
min_vals = 1000000 * min_vals
avg_vals = avg_vals.tolist()
max_vals = max_vals.tolist()
min_vals = min_vals.tolist()

# if baseline
if len(new_units) == 2:
avg_vals[1] = str(avg_vals[1]) + " " + str(avg_percent_diff)
max_vals[1] = str(max_vals[1]) + " " + str(max_percent_diff)
min_vals[1] = str(min_vals[1]) + " " + str(min_percent_diff)

df.loc[df["Metric"] == curr_metric, "Avg"] = avg_vals
df.loc[df["Metric"] == curr_metric, "Max"] = max_vals
df.loc[df["Metric"] == curr_metric, "Min"] = min_vals
df.loc[df["Metric"] == curr_metric, "Unit"] = new_units[0]
if "Metric" in df:
curr_row = df[df["Metric"] == curr_metric]
if not curr_row.empty:
# fix values
if "Value" in curr_row:
vals = curr_row["Value"].values
new_units = []
percent_diff = 0

# if baseline
if isinstance(vals[0], np.ndarray):
val_1 = curr_row["Value"].values[0][0]
baseline = curr_row["Value"].values[0][1].split()
val_2 = float(baseline[0])
percent_diff = baseline[1]
vals = np.array([val_1, val_2])

# calculate units
for val in vals:
if isinstance(
val,
float,
):
if val < 0.001:
new_units.append("Kb/s")
elif val < 1:
new_units.append("Mb/s")
if len(new_units) == 2:
if new_units[0] == "Kb/s":
new_units[0] = "Mb/s"
else:
new_units.append("Gb/s")
if len(new_units) == 2:
new_units[0] = "Gb/s"
# Convert to new_units
if new_units[0] == "Mb/s":
vals = 1000 * vals
elif new_units[0] == "Kb/s":
vals = 1000000 * vals
vals = vals.tolist()
# if baseline
if len(new_units) == 2:
vals[1] = str(vals[1]) + " " + str(percent_diff)

df.loc[df["Metric"] == curr_metric, "Value"] = vals
df.loc[df["Metric"] == curr_metric, "Unit"] = new_units[0]

elif "Avg" in curr_row:
avg_vals = curr_row["Avg"].values
max_vals = curr_row["Max"].values
min_vals = curr_row["Min"].values
new_units = []
percent_diff = 0

# if baseline
if isinstance(avg_vals[0], np.ndarray):
avg_baseline = curr_row["Avg"].values[0][1].split()
avg_percent_diff = avg_baseline[1]
avg_val_1 = curr_row["Avg"].values[0][0]
avg_val_2 = float(avg_baseline[0])
avg_vals = np.array([avg_val_1, avg_val_2])

min_baseline = curr_row["Min"].values[0][1].split()
min_percent_diff = min_baseline[1]
min_val_1 = curr_row["Min"].values[0][0]
min_val_2 = float(min_baseline[0])
min_vals = np.array([min_val_1, min_val_2])

max_baseline = curr_row["Max"].values[0][1].split()
max_percent_diff = max_baseline[1]
max_val_1 = curr_row["Max"].values[0][0]
max_val_2 = float(max_baseline[0])
max_vals = np.array([max_val_1, max_val_2])

# calculate units
for val in avg_vals:
if isinstance(
val,
float,
):
if val < 0.001:
new_units.append("Kb/s")
elif val < 1:
new_units.append("Mb/s")
if len(new_units) == 2:
if new_units[0] == "Kb/s":
new_units[0] = "Mb/s"
else:
new_units.append("Gb/s")
if len(new_units) == 2:
new_units[0] = "Gb/s"

# Convert to new_units
if new_units[0] == "Mb/s":
avg_vals = 1000 * avg_vals
max_vals = 1000 * max_vals
min_vals = 1000 * min_vals
elif new_units[0] == "Kb/s":
avg_vals = 1000000 * avg_vals
max_vals = 1000000 * max_vals
min_vals = 1000000 * min_vals
avg_vals = avg_vals.tolist()
max_vals = max_vals.tolist()
min_vals = min_vals.tolist()

# if baseline
if len(new_units) == 2:
avg_vals[1] = str(avg_vals[1]) + " " + str(avg_percent_diff)
max_vals[1] = str(max_vals[1]) + " " + str(max_percent_diff)
min_vals[1] = str(min_vals[1]) + " " + str(min_percent_diff)

df.loc[df["Metric"] == curr_metric, "Avg"] = avg_vals
df.loc[df["Metric"] == curr_metric, "Max"] = max_vals
df.loc[df["Metric"] == curr_metric, "Min"] = min_vals
df.loc[df["Metric"] == curr_metric, "Unit"] = new_units[0]
return df


Expand Down

0 comments on commit 75c7b25

Please sign in to comment.